Raytheon Opamp?

[Bassmankr]

Just stripped an AMR/Peavey 2400 console I picked up and I see this opamp in the modules (top line RA3678C, bottom line R8924) which I'm assuming is a Raytheon RA3678C. Its in sockets that call for a National 5532 (specifically in the manual/partslist a NE5532FE Dual Lo-Noise). I can't find a data sheet on line and my digikey catalog doesn't carry Raytheon. My question finally, is this something someone did as an upgrade or is this a generic/low cost part for a 5532 that AMR/Peavey put in the desk? This desk also has lots of TLO-72 Duel bifet JG PKG in the form of Texas Instruments TLO72CP (8 dip in sockets) and a few TLO-74 quad bifet in the form of Texas Instruments TLO74C. Are there reasonable upgrades that I could just pop in the sockets? By the way, the TLO72C's in the input modules have a silver color top where the TLO72C's in the duel moniter modules are all black (same stencil/numbering/logo on both), does the silver top color mean anything? Either AMR put in some nice parts or a prior owner upgraded the caps since they are Nichion so I'm guessing I won't need to do anything to the caps till they go, or am I wrong to assume this?

[JohnRoberts]

Originally Posted by Bassmankr

Just stripped an AMR/Peavey 2400 console I picked up and I see this opamp in the modules (top line RA3678C, bottom line R8924) which I'm assuming is a Raytheon RA3678C. Its in sockets that call for a National 5532 (specifically in the manual/partslist a NE5532FE Dual Lo-Noise). I can't find a data sheet on line and my digikey catalog doesn't carry Raytheon. My question finally, is this something someone did as an upgrade or is this a generic/low cost part for a 5532 that AMR/Peavey put in the desk? This desk also has lots of TLO-72 Duel bifet JG PKG in the form of Texas Instruments TLO72CP (8 dip in sockets) and a few TLO-74 quad bifet in the form of Texas Instruments TLO74C. Are there reasonable upgrades that I could just pop in the sockets? By the way, the TLO72C's in the input modules have a silver color top where the TLO72C's in the duel moniter modules are all black (same stencil/numbering/logo on both), does the silver top color mean anything? Either AMR put in some nice parts or a prior owner upgraded the caps since they are Nichion so I'm guessing I won't need to do anything to the caps till they go, or am I wrong to assume this?

The 3678 is actually a custom in-house PN for a 5532 that is selected for noise floor (8924 is a date code). The noise floor spec was a bit unusual in that it wasn't an absolute level spec but a contour or spectral balance of the noise floor. This grading would weed out any parts with 1/F (process related) noise problems and guaranteed a smoother sounding noise floor.

In the sockets where they were used mostly as output drivers you can drop in a 5532 or equivalent if you need to. I think the only place they were used where noise is even remotely an issue might be in the sub master summing bus amps, but unless you have tens of inputs assigned to one sub, the noise gain will be modest and again a standard 5532 could be substituted.

Note: replacing with a 5532 is not an upgrade since the 3678 is already a selected 5532. You can't buy them from Raytheon, but you may be able to still get them from Peavey parts dept if they still haven any inventory on them.

The color of the bi-fets has no significance, they were whatever color TI felt like making them at the time. Nichicon was one of several approved cap vendors so they may be original.

JR

PS: If you have a full load (36x24) and you're thinking about serious opamp upgrades you may want to do some back of the envelope math regarding PS current since there's more than a couple in there.

[Jim Williams]

Try the new National LM4562NA in the stereo sum position. It's much less noise, distortion, more slew rate/bandwidth.

Jim Williams
Audio Upgrades

[JohnRoberts]

Originally Posted by Jim Williams

Try the new National LM4562NA in the stereo sum position. It's much less noise, distortion, more slew rate/bandwidth.

Jim Williams
Audio Upgrades

That looks like a nice part on paper and probably probably improves on some off-sheet aspects of the 5534 too. That old part is a little long in the tooth. I didn't find a current consumption spec from a quick google, but they should drop right into the sub master summing sockets. I wouldn't expect a night and day sonic difference in the subs, but any perceived difference will likely be in the direction of improvement.

My only caution regarding the master L/R socket is that some versions of that console used a hybrid (discrete low noise bipolar transistor with opamp wrapper). If your console is using just a bare 3678 this sounds like a worthwhile upgrade. If not the input noise will still be dominated by the transistor (IIRC 2SD786 @ .5 nV/rt Hz or so) and you would need to confirm that the circuit is still stable with a scope. The extended GBW can work against stability when used inside a feedback loop.

That 55 Mhz GBW product is much improved over 3678/5532 technology so it should deliver lower phase shift and improved distortion at 20kHz when working at high noise gains.

Have fun it's only money, but do try a little harder than I did to find out PS current consumption before swapping out on more than just bus summing amps. I don't know how much PS headroom was left over in the 36x24 full load.

JR

[Bassmankr]

Thanks for the info guys,

John, I do have the full load config on the desk 36 left side input modules and 24 right side duel moniter modules so I'll be very cautious with the math and mods to make sure I don't stress/overload the PS. I came across that right side of the desk panning trick when I was researching whether to buy the desk or not but thanks for the tip regardless. Lou Gimenez in NY is a frequent poster here and has the 16 buss version of this desk, he gave me some contact info for a tech in NY named Tom Maguire who did a mod of his master module. What he did exactly to it Lou didn't say but he told me it certainly "opened the desk up" alot. I'll be calling him soon but I wanted to learn what makes this desk tick so I could have an intelligent conversation with him (I'm having a blast and really learning alot). As for the sound of the desk before I stripped it to the frame, the mids and lows were right there and creamy but the high end was veiled (maybe thats what Lou meant when he said the mod "opened it up". I could make up for it with the EQ but that's not the point, you want the desk sounding great without the EQ kicked in.

Jim warned me that the construction is bad in these desks and he was right but some of that is an easy fix. Layout on the input modules are bad also in that the EQ is up near the meters and buss assigment switches are close to the faders so you have to do lots of reaching, it would be much better if these two items switched positions. The last big negative is the real estate the desk takes up, 8 feet wide LOL but given the song I paid for it and the fact that it's replacing a Mackie 32-8 plus the learning experience of a desk that I can work on makes me think I made a compromise in the right direction.

[JohnRoberts]

Originally Posted by Bassmankr

Thanks for the info guys,

John, I do have the full load config on the desk 36 left side input modules and 24 right side duel moniter modules so I'll be very cautious with the math and mods to make sure I don't stress/overload the PS. I came across that right side of the desk panning trick when I was researching whether to buy the desk or not but thanks for the tip regardless. Lou Gimenez in NY is a frequent poster here and has the 16 buss version of this desk, he gave me some contact info for a tech in NY named Tom Maguire who did a mod of his master module. What he did exactly to it Lou didn't say but he told me it certainly "opened the desk up" alot. I'll be calling him soon but I wanted to learn what makes this desk tick so I could have an intelligent conversation with him (I'm having a blast and really learning alot). As for the sound of the desk before I stripped it to the frame, the mids and lows were right there and creamy but the high end was veiled (maybe thats what Lou meant when he said the mod "opened it up". I could make up for it with the EQ but that's not the point, you want the desk sounding great without the EQ kicked in.

Jim warned me that the construction is bad in these desks and he was right but some of that is an easy fix. Layout on the input modules are bad also in that the EQ is up near the meters and buss assigment switches are close to the faders so you have to do lots of reaching, it would be much better if these two items switched positions. The last big negative is the real estate the desk takes up, 8 feet wide LOL but given the song I paid for it and the fact that it's replacing a Mackie 32-8 plus the learning experience of a desk that I can work on makes me think I made a compromise in the right direction.

Perhaps ironic IMO it was Mackie's 8-bus in-line for just under $4k retail that killed the market for the Peavey $10k split... Just look at the price points now for value 8-bus desks. Hundreds not thousands of dollars. In hindsight Peavey/AMR may not have been the obvious brand to market an old school split into an increasingly in-line market.

I'll try not to sound like I'm making excuses but my layout guys were tasked with using in house single-sided PCB technology with crude by today's standards (20+20) spaces and traces. Physical layout was somewhat influenced by natural signal flow and the difficulty in passing signals back and forth. Even with modern fine line technology short signal paths are better. Short is a relative concept in the context of 6’ long consoles. As i recall I had next generation module revs working on the bench but the market went away from us pretty quickly, and I had to retarget my engineering group’s resources elsewhere.

I'll be glad to share my $.02 off list on specific tweaks and try not to get too defensive (I am human with a healthy ego). I think I recognize Tom Maguire, so I have probably spoken with him before. There were several AMR consoles in NYC area.

Good luck and with 100+ feeds to the L/R bus you're not likely to run out of inputs during mix down.

JR

[Bassmankr]

John, didn't know you had something to do with the AMR desk, I'd love to have a chat with you. I assume I can just email you at your website to reach you. Hope I didn't offend when I was talking about the ergonomics of the desk. I will miss the compact inline design but sonics mean so much more and thats why I'm going this route. I too see the irony but what is even more ironic is the swing back to analog desks in computer based studios. It seems to me that more guys like hybred setups using the strengths of both analog and DAW. I'm thinking that demand is going to go up in the near future for used analog desks with decent sonics and that I bought at the right time (low part of the cycle).

[JohnRoberts]

Originally Posted by Bassmankr

John, didn't know you had something to do with the AMR desk, I'd love to have a chat with you. I assume I can just email you at your website to reach you. Hope I didn't offend when I was talking about the ergonomics of the desk. I will miss the compact inline design but sonics mean so much more and thats why I'm going this route. I too see the irony but what is even more ironic is the swing back to analog desks in computer based studios. It seems to me that more guys like hybred setups using the strengths of both analog and DAW. I'm thinking that demand is going to go up in the near future for used analog desks with decent sonics and that I bought at the right time (low part of the cycle).

No worries, I'm not easily offended especially about stuff. I worked at Peavey for 15 years and as you can imagine I didn't get much love from gear snobs.

Yes, you can contact me directly via my web domain ( john_roberts@ ). I don't have any engineering documentation but do recall the basic design philosophy.

JR

[Tiz]

Good Stuff, JR! I own an AMR 1600 board and would love to get some tips well. Coincidentally I just got in some of those new National LM4562NA, I think I'll try them out in the L/R sum bus as Jim suggests. Bassmankr, send me your email, it'd be good to get to know another AMR owner.

[JohnRoberts]

Originally Posted by Tiz

Good Stuff, JR! I own an AMR 1600 board and would love to get some tips well. Coincidentally I just got in some of those new National LM4562NA, I think I'll try them out in the L/R sum bus as Jim suggests. Bassmankr, send me your email, it'd be good to get to know another AMR owner.

As I mentioned earlier in this thread some versions of that console series used discrete devices in the L/R summer so dropping in a new opamp may offer little or no benefit..

I was pleased with the input strip EQ on those desks but of course I'm biased.

JR

[Bassmankr]

It would be great if we could pool all the AMR desk owners with a contact list for info on mods and stuff, you can reach me at bassmankr@ and the domain is fvi.net (listed like this to cut down on some of the spam)

[Jim Williams]

Originally Posted by JohnRoberts

As I mentioned earlier in this thread some versions of that console series used discrete devices in the L/R summer so dropping in a new opamp may offer little or no benefit..

I was pleased with the input strip EQ on those desks but of course I'm biased.

JR

The mix opamp will make a massive difference , even with the transistor front end. Replace those slow 2SD transistors with Hitachi 2SC2545's, it's like your ears popping after a long flight. .5 nv/hz/sq noise spec and 100meg hz ft's work much better. For even lower noise, pop in a new AD8599 dual with it's 1 nv/hz noise spec. Keep in mind resistors in the circuit may dominate the noise spec, replacing those with Dale RN55's may also help. Sometimes a wideband opamp loses phase margin with these transistor grounded base front ends. A small cap of around 220~330 pf fitted between the opamp's input pins raises the noise gain of the opamp and restores the phase margin = stability. Also add close by local .1 uf caps to the power supply and ground.

Noise is one spec, but massive slew rates offer other benefits such as lower crossover distortion due to faster switching output devices on the opamp. With the cap suggestions do try out a National LM6172 and the Linear Tech LT1358. Both of these offer a more open top end than the LM4562 or AD8599.

Jim Williams
Audio Upgrades

[JohnRoberts]

Originally Posted by Jim Williams

The mix opamp will make a massive difference , even with the transistor front end. Replace those slow 2SD transistors with Hitachi 2SC2545's, it's like your ears popping after a long flight. .5 nv/hz/sq noise spec and 100meg hz ft's work much better. For even lower noise, pop in a new AD8599 dual with it's 1 nv/hz noise spec. Keep in mind resistors in the circuit may dominate the noise spec, replacing those with Dale RN55's may also help. Sometimes a wideband opamp loses phase margin with these transistor grounded base front ends. A small cap of around 220~330 pf fitted between the opamp's input pins raises the noise gain of the opamp and restores the phase margin = stability. Also add close by local .1 uf caps to the power supply and ground.

Noise is one spec, but massive slew rates offer other benefits such as lower crossover distortion due to faster switching output devices on the opamp. With the cap suggestions do try out a National LM6172 and the Linear Tech LT1358. Both of these offer a more open top end than the LM4562 or AD8599.

Jim Williams
Audio Upgrades

My original response to this was lost when the board was down the other day. I must have had bad karma that day.

JR

[Bassmankr]

I read it and wished I had made a copy to a text file, sad that it's lost. I know its a pain to remember and repost, I've lost a few of my long posts on forums and its a pain.

[JohnRoberts]

Originally Posted by Bassmankr

I read it and wished I had made a copy to a text file, sad that it's lost. I know its a pain to remember and repost, I've lost a few of my long posts on forums and its a pain.

Maybe this weekend.. I recall most of what I said, still a PIA to do anything twice.

I'm new here does this happen often?

JR

[Bassmankr]

its a fairly stable forum compared to others but when I get log winded I right click copy my post because it usually logs me off and that's where I've lost a few posts. Since I copied the post I can always just hit the back button a few times and just redo my reply with a fast right click paste.

[JohnRoberts]

Originally Posted by Bassmankr

I read it and wished I had made a copy to a text file, sad that it's lost. I know its a pain to remember and repost, I've lost a few of my long posts on forums and its a pain.

OK, here a revisit of my post but paraphrased:

As I recall I was responding directly to Jim Williams who suggested replacing even the discrete summing amps "will make a massive difference" (his words). I don't agree but it's hard to argue a hypothetical and I may have more than a little personal bias and ego..

In my experience summing amps in large bus structures (the AMR with full load has as many as 112 separate sources feeding the L/R on mix down) will be more audibly impacted by things like loop gain margin on linearity and phase response than noise.

Even just a handful of channels wide open will swamp the bus noise floor with the combined noise of their microphone's self noise amplified by even 50-60 dB of noiseless gain, not to mention amplified room ambient.

OTOH the summing amp when operated at high noise gain will be at the mercy of reduced loop gain margin (difference between open loop and closed loop gain). This gain margin factor impacts the opamps ability to linearize open loop distortion, and correct for the open loop transfer function's 90' of phase shift due to internal stability compensation.

Another data point is that the AMR summing bus topology is not typical in that the left side of the console uses conventional resistors to feed the summing bus, but the right side (monitors) use synthesized current sources on the three feeds per strip over there. This dramatically reduces the accumulated noise gain at the summing bus that usually consumes available loop gain margin. In the worst case example of 112 active feeds to mixdown the actual noise gain of the AMR is something like 40x. Further, and I don't think I mentioned this in my first post, there is 10 dB of cut available in the feedback network of the summing amp which will reduce that noise gain by up to another 3x.

So if you have the time and money try a swap, I would just temper your expectation about realizing "massive differences".

======
Jim also made a suggestion about resistors "Keep in mind resistors in the circuit may dominate the noise spec, replacing those with Dale RN55's may also help."

In general the noise in resistors is caused by random currents due to temperature (Johnson Noise) and more a factor of value than type. There is a mechanism called "excess noise" which is related to uniformity of the resistive substrate and modulated by current but this is primarily a consideration in vacuum tube bias circuits and such where there are large DC biases present. Modulation noise below audio current is too far down to perceive due to masking. This is a perhaps valid (in theory) consideration for dropping resistors in mic phantom voltage circuitry which already typically use MF. Finally there can be nonlinear voltage coefficients and such related to cheap resistors (I've heard about distortion in much higher voltage swing power amp feedback networks), but to my knowledge the AMR consoles used quite decent carbon film in all audio paths. IIRC the AMR used relatively low impedance summing resistors to keep impedances low to manage crosstalk between buses, so Johnson noise there should not be an issue.
------------
Lastly I think I asked Jim for his opinion about a new mixed process (Si & Ge) opamp I had heard about http://www.ti.com/corp/docs/landing/opa211/ . I do not follow the latest opamp du jour so don't know if this is even news or how they might compare to his short list of favorites.

______

I think this is pretty much all I posted about.

FWIW back in the '80s I wrote an article on console design published in R E/P magazine (now defunct). In that article I described the current source summing topology and a few other design issues. I didn't discuss the AMR console as that was designed a decade or so later .

JR

PS: If this gets erased again it was meant to be...

[Jim Williams]

First off, folks here should know that John is a great designer and one who isn't afraid to incorporate new design techniques and topologies. His articles "Performance Limitations for Contemporary Console Design" in 1980 for the defunct REP magazine are still (I saved some of these) some of the best design topology discusssions one can read on the subject. No, we're not talking about old Neve summing amps or even any of the new summing boxes but rather the use of discrete transistors and opamps to achieve near theoretical specs on noise and distortion.

These topologies for mic preamp design were called "Trans-Amps" (credit due to Paul Buff) and offered better specs with transistors than with transformers. The summing amp version, commonly known as grounded base, offers incredible dynamic range and bandwidth not ahievable with common opamp inverting summing designs. Besides those versions of AMR, this design or it's variants is used in modern Soundcraft, D+R, some Neotek, and a version with base inputs is used in SSL's. I have measured -90 db stereo crosstalk @ 10k hz on the Soundcraft I use and a breadboard version I tested with the AP measured -134 db THD + noise referenced to +4 dbu. Put in some high level signal at around +15 dbu and it's 10 db better. This is why I can't use digital summing, it's not as good.
As most choose digital summing for it's clarity, I use this high end analog summing for the same reason.

With these specs, John is correct in that virtually all noise and grit will be sourced from the feeds to the summing amp, the inputs. Still, it's nice to know the mix stage is never going to be a problem. I have noticed a sense of extra clarity and openess when I replaced the 2% Asian metal film resistors with the Dale RN55's and RN60's. I would suspect the difference would be even more if the carbon films were exchanged. Then again, I went as far as to remove the stereo conductive plastic faders and it's Kimber silver teflon wiring. I have spare modules with the same parts/mods but with the original resistors to compare to and there is a difference.

I've been playing with both the OPA211 and the fet input OPA827. Both are fine amps. I also have some of the AD8599's which are duals like the LM4562. The AD is quieter, but the National is lower dirt and faster. The fet OPA827 is a great sub for the OPA627 which is overpriced and sonically bland. I'm evaluating these for a couple of design projects I'm working on in the MI field so they are going to be selected for sonics, not specs. The lower noise floor of these chips allows for lower noise designs so noise isn't involved in selection, just the tone.

All of these new opamps are missing the top end open clarity of some of the better wideband opamps. Although some of these new chips have 80 db open loop gain in the audio band, the wideband parts with high slew rates and 80 db OLG at 10k hz still sound better. As clear as these new chips are, there is still something missing at very low levels in the high frequencies, commonly heard with percussion transients. As the test gear doesn't show these differences, it becomes a listening test.

I had a customer that had a Audio Upgrades mic pre in for repair after a damaged mic was used. I installed the chips I liked as they are the clearest but he called and said he wanted the older parts re-installed. He said in close miking he had too many details to work with, for distance or orchestral work it would be great. So I sent more of the older parts to him and let him keep the new ones in case of an acoustic distance recording.

I'm now down to selecting chips for tone and applications. The AP isn't helping at all. On customer in DC uses ProStools 192 into a modified Harrison MR-4 with Uptown faders. It has custom "grounded base" summing cards I installed as this is 2 56 input consoles bolted together. Grounded base was the only way to sum 112 inputs without the summing opamp running out of loop gain. Same situation as before, I sent some opamps to install in the summing cards. They called and said it sounded amazing but they were hearing too many ProStool "artifacts". Re-installing the older, darker chips seemed to fatten up the sound while dampening the digital "uglies"

Jim Williams
Audio Upgrades

[Jim Williams]

One other thing. The AMR uses a unique EQ circuit. I'm going by memory (I don't want to dig into the filing cab) but I believe it uses an all pass filter function that avoids running the audio signal directly through any EQ capacitors. Ben Duncan also wrote an article in Studio Sound discussing this new and interesting topology.
Maybe John can chime in and offer some insights?

Jim Williams
Audio Upgrades

[Wavebourn]

John;


out of curiosity, why don't split summing between several summing amps, say 10 amps with 12 inputs each?

--Anatoliy

[Jim Williams]

John covered that technique in the REP articles. It's a form of distributed bussing when inputs are grouped to sub summing amps. It buys about a 3db improvement but is nothing like the trans-amp grounded base design. If you feed the grounded base design with current sources, the number of inputs is almost unlimited. Any inverting opamp summing amp will be restricted by it's loop gain. This is why a few inputs mixed on a Trident 80 sound OK, but when you bring up 40 faders, it turns to mud.

Jim Williams
Audio Upgrades

[JohnRoberts]

Originally Posted by Jim Williams

One other thing. The AMR uses a unique EQ circuit. I'm going by memory (I don't want to dig into the filing cab) but I believe it uses an all pass filter function that avoids running the audio signal directly through any EQ capacitors. Ben Duncan also wrote an article in Studio Sound discussing this new and interesting topology.
Maybe John can chime in and offer some insights?

Jim Williams
Audio Upgrades

No the EQ in the AMR was not novel, but it was nice. It was a wein-bridge variant which I found attractive since the topology is subtractive only (in the input path for cut, feedback path for boost). This results in much less wide band noise gain on the opamps involve compared to a typical bandpass section with it's output noise boosted by the max boost/cut of the EQ section.

I can't draw the schematic from memory, that was a couple of decades ago and I nicked it from an engineer in Peavey's mixer group, who probably nicked it from somebody else. I didn't ask, it looked quieter and cleaner than my prior work (state variable parametric) which wasn't in the budget for this series anyhow (not enough room for all the knob) so I dialed it in for a Q more appropriate for studio use.

The only downside to that particular topology which as I recall was only one opamp, was that there was a strict relationship between the Q and max boost/cut available, but those characteristics worked out to a nice range for the application so it was easy to ignore.

JR

PS: Sorry I don't have schematics but a lot of water has flowed over that dam. There were schematics published in both the AMR recording console and PV Mark VIII SR console owner's manuals so I'm sure somebody out there has them.

PPS: wrt Wavebourns question, the 3db was for every doubling of the number of summing amps so you quickly reach dininishing returns, at least wrt noise floor, which IMO isn't the big nut anyhow.

[JohnRoberts]

Originally Posted by Jim Williams

I have noticed a sense of extra clarity and openess when I replaced the 2% Asian metal film resistors with the Dale RN55's and RN60's. I would suspect the difference would be even more if the carbon films were exchanged.

Jim Williams
Audio Upgrades

First, thanks for the kind words. That old article holds up pretty well, but we've just been adding zero's on the right side of the decimal point for decades now. I hope I don't disappoint any with expectations of huge improvements from here. The modern SOTA is pretty damn good.

I am guilty of generalizations about resistors. For "excess noise" WW was supposed to be best and MF was second, but I recall a white paper maybe circa mid '70s about a specific manufacturer's series of carbon comp that beat MF in measurements. Of course this also gets complicated with 1/2 W vs. 1/10th W etc. (since excess noise is related to uniformity of resistive substrate it matters). We can perhaps be more confident in talking about what we know is good, than what we assume is bad.

I have had very good experience with "Asian" CF but they happened to be ROHM (Japanese) parts I used since the '70s. They were IMO unusually good. For fun we would heat them up till red hot and then after they stopped glowing measure and find many still within their 5% original tolerance band, of course their resistance value markings were now smoke. Of course this doesn't prove anything other than that you shouldn't use them as fuses , but I had nothing but good experiences with them. ROHM was already the approved resistor vendor at Peavey when I got there and AFAIK were the parts used in the AMR consoles.

They were later second sourced by a cheaper vendor (over my objections and after a rigourous production trial). It was hard to fight the kind of dollar savings even a few tenths of a cent can add up to across a few bazillion parts per year. The unadvoidable double edged sword of piggybacking on the tails of a manufacturing giant.

Back to your point, I have nothing but good experiences with mil-spec RN55 MF parts, I just would not expect an audible improvement over ROHM CF. YMMV

FWIW I have seen a resistor actually cause measurable distortion in a power amp feedback network so your caution is justified. Like I always say to hot-rodders, "It's your money and your time so just be sure to do no harm" swapping a good CF for a good MF will not cause any harm.

It might be interesting for somebody with adequate extra time on their hands and a suitable test bench to test these audibly bad resistors for linearity..? if you can hear a difference surely there must be a measurable difference. Of course If resistors are suspect, everything is suspect. Uh oh...

Merry Christmas

JR

[Wavebourn]

Instead of measurements I couple of years ago found a seller on ePay and bought a huge lot of Soviet military MLT-05 resistors (such red metal films with silvered legs). I don't remember his ID, his first name is Kyle, he is from some Baltic country.

[Bassmankr]

John, after the holidays when I get some more time I'd be glad to make you a copy of my 2400 manual/schematics and send it down to you.