Altec 9440A Recap

[miscrms00]

Just as I finished getting my Altec 1608As repaired, recapped, and ready to go back in the rack plans changed.
http://audiokarma.org/forums/index.php?threads/any-hope-for-this-altec-1608a.683434/

I came across a pair of Altec 9440As at a good price on ebay, and the seller turned out to be just down the street. I had been missing the one I traded off a few years ago, so it seemed like an opportunity not to be passed up. And there was no doubt these pack a lot more punch into a lot less rack space than 1608As, even if they do lack the iconic green fronts. These were the last two of a dozen or more that had been used in his professional sound event company, but had been long since retired for most events by newer more powerful amps. Funny thing, when I got there I realized he was the one whose stickers had been on some of the retired Altec/JBL gear I had previously acquired from ASU, as he had been the local dealer / service center for Altec and JBL for many years. Pretty cool! I was not disappointed when I picked them up, aside from some minor scuffs / blemishes they were very clean and look to have been well maintained as well as very close in serial numbers. DC offsets were not very far out, but I was able to dial them both in to about 1mV or so.

I've already had the opportunity to do some stress testing on these amps in the course of my recent Intersonics ServoDrive Contrabass rebuild, and they seem to be working like a champ even with sustained continuous sine waves at 100% output into a 3-4 Ohm load.
http://audiokarma.org/forums/index....servodrive-contrabass-rebuild-attempt.754517/

Never the less, it seemed like it would be a smart idea to go ahead and replace the electrolytic caps before putting them into daily service, as the owner couldn't really say when they'd last been done due to their semi-retirement. So here's what I've ordered. I decided to go mostly with a Vishay series similar to what I used in the 1608As which come from a legacy Philips plant in Austria. Although this time I decided to step up to the 138 AML series from the 021 series to get 105C temp rating. They were not cheap, but I wanted to stick with axial and with what digikey has in stock. In some cases I have gone higher in voltage as these are physically smaller than the originals, often have better specs, and are sometimes cheaper for some reason.

Control PCB:
C2 10uF 50V: Vishay 138 AML 10uF 100V
C3 200uF 25V (radial): Kemet ESY 220uF 50V
C5 100uF 50V: Vishay 138 AML 100uF 50V

Driver PCB (x2):
C1 5uF 25V: Vishay 138 AML 4.7uF 63V
C3 50uF 16V: Vishay 138 AML 47uF 40V
C5 10uF 25V: Vishay 138 AML 10uF 25V (ordered extra 100V too if they fit).
C10/11 10uF 100V: Vishay 138 AMl 10uF 100V

Chassis:
C1/2: 9800uF 100V: TDK EPCOS SIKOREL (Siemens) B41550 10,000uF 100V

The pair of main PS caps were pretty painful, but I figured I want to keep these amps a long time and the 20,000hrs at 105C sounded pretty good. Last time I did a 9440A I used the B41456 line, half the cost but "only" 12,000hrs at 85C. They are also only 2" diameter so they took some finagling to fit in the 2.5" holders. The B41550 are proper 2.5". All in all about $220 worth of caps from digikey, not terrible I suppose for 2 amps. Actually less per channel than I spent on the 1608As if you want to look at it that way.

So anyway, once the caps arrive I'll plan to trace things out and check resistors, voltage, etc along the way to see if anything else looks suspect, but I don't expect to find much given how well they seem to be working.

Here's a few pics of the internals to get rolling:

 

[miscrms00]

And schematics for anyone that wants to play along at home...

 

[macyjrm]

Watching this post with considerable interest. I know where a pair of these exist no longer in use but haven't yet been able to acquire them. Still hopeful.

 

[miscrms00]

Caps arrived today! Nice to see that the 138 AML series Vishay caps were also all labeled with either Austria or Denmark as country of origin like the 021 ASM series. The white ones are even still labeled BC I believe.

I decided to try making a before / after noise plot to see if there is any appreciable difference after changing out the caps. I figure as long as the mic and speaker are in the same position it should be a decent comparison. UMIK-1 is placed right in front of speaker (JBL S38) on channel 1, turned up to 10. Should have done channel 2 as well, but didn't think of it until the amp was apart. Blue curve is
with the amp off to have a reference of the noise in the room.

Removed the big PS caps, and then the driver board on the right side.

Nice to have a board with all component locations and cap polarities already marked! All resistors ohmed out in spec, all diode/transistor junctions seem ok.

After swapping out all electrolytics. The old ones all tested ok on the 4192A, but some values had increased by 20-50%, indicating they were showing their age.

It may be worth replacing the input coupling cap C1 with something better at some point. Particularly as I failed to notice this one was radial and ordered an axial.

Rob

 

[miscrms00]

Finished the first amp, seems to be working well. Can't say for sure that there is much difference, but they didn't particularly have any issues before. Just figured it would be good to replace the electrolytics before putting into daily use. Its possible the bass response is somewhat improved, as I started with the treble and bass settings on the pre-amp set flat and didn't find myself immediately reaching to turn up the bass as I usually do I didn't end up doing the front control boards, as they seemed to be a royal pain to get out. Hopefully that won't end up being an issue down the road.

Second driver board before and after, looking very similar to the first.

I just picked up batch of NOS Siemens polypropylene and polystyrene caps, it might be interesting to try replacing the ceramics with those at some point. Seems like the board is pretty roomy, so probably wouldn't take too much shoe horning. Or to try putting one in parallel with the input cap.

I did note some differences in schematics and the boards. The paper copy I have from the seller the driver board schematic is marked -06, while the electronic version attached is marked -08. The boards themselves are marked -10, though using a different part numbering system so not 100% sure the rev numbers are consistent. In the -06 schematic R5 and R10 are 22k 1/4 watt. In the -08 schematic they are marked 6.8k 2W, same as R13. The boards appear to match the -08 schematic, they are the two big resistors at top left in the images above.

One other note, in checking out the resistors on the board, all but 3 were measuring in spec in situ. On the first board I removed one leg on each of those 3 and verified that they measured in spec once disconnected. On the second board I just verified that they read the same as the first board in place. The 3 were R36 (measured 5.1k instead of 10k), and R39/40 (measured 3.8k instead of 5.1k). Also in measuring junctions all but CR 17/18 seemed reasonable. These ones measured 0.28V and were symmetrical with polarity. Noting in the schematic that they are in parallel with a 180 Ohm resistor this seems to make sense.

The Sikorels fit well into the holders, but are about an inch shorter. This doesn't seem to be an issue, as there is enough slack in the wiring to reach. They actually make is a little easier to access the screws when re-installing. I probably could have used the 15,000uF Sikorels, as I believe they would have been back to stock height. Might have been worth trying but they were quite a bit more expensive as I recall, and i was already smarting a bit from buying 4 of these.

IMPORTANT NOTE: Both main PS decoupling caps are installed with their + side facing up. The bottom is -80 to GND (+), and the top is GND to +80 (+). Somehow I did a very dumb thing and got the top one in backwards. I think I just really wanted to see the pretty gold label I should have checked more carefully, and should have remembered this from the last one I did. I made a very dumb assumption instead of checking directly. The second amp that was still together had the caps installed label side up. The caps that came out of the first amp looked to be identical but have their label near the - terminal so they would have been in label side down. The new caps have their label near the - terminal, same as the old ones sitting next to them. And the picture I was referencing was of the second amp, not the first. So I initially installed the top cap with polarity reversed. Fortunately I noticed something didn't seem right when I powered it up. Voltage measured +50V instead of +80, and the top cap started getting warm. Luckily I seem to have caught it before any damage was done. After pulling it out and letting it cool it still measured the same as the others in terms of C and D. Hopefully that won't turn out to be a $30 brain fart.

I try to re-measure the noise, but realized I had the amp hooked up wrong when I did the baseline. So a direct before/after comparison didn't end up being possible. I may try again on the second amp, and hopefully do a better job. Still not sure its worth much

Ch 2 may be a bit nosier at low frequencies, channel 1 a bit more at high frequencies. All were quiet enough that I couldn't hear anything without getting my ear right by the speaker, though the room is not super quiet. I also don't plan to run these with the volume at 10 though

Power supply voltages are still measuring pretty much spot on +/- 80V. I may try to check some of the internal voltages but they are difficult to access when connected. I had previously set the output DC offset down from 10-20mV to around 1mV on each channel. Should check to see if they are still there. Probably should have hit the open pots on the driver boards with some de-oxit. Have not yet checked the output bias adjustments, and not sure I'd trust my cheap ammeter anyway.

 

[Moving Ahead]

Looks like you'll have to reinforce the floor under your rack to accommodate all this heavy metal.

Heres a few notes that you can take or leave.
The larger axial caps look better. if theres a lot of lead bling I'll use some spaghetti to hide that. But yours looks pretty good.
Have you considered replacement sealed trim pots?
How matched are Q1 & Q3 do you think?
There's a fair amount of ceramic in there.
C1 is small enough to put stacked film.

Look forward to seeing your finished work.

 

[miscrms00]

Good question on Q1/3. Since thats's the input diff pair I assume that's pretty key. After a fair amount of cursing and knuckle bashing I was able to sneak some micro grabbers in and measure the Vbe on each channel. Measured with no input, here's what the pairs look like:

581 / 572 mV : 9mV dc offset

579 / 587 mV : 8mV dc offset

That seems decent to me in the context of input swing, which I think is something like 1.4Vpp? What do you think?

Agreed on the ceramics. I don't know how much difference it will make but its probably worth doing, particularly if I have something in my stash that will work. I like the idea of replacing the pots too... They do look a bit crudy,

Any thoughts on what a reasonably priced improvement to electrolytic might be for the 5uF input coupling cap? I think the biggest I have in film is 3.3u mylar. My PS and PP are all below 1uF I think.

 

[miscrms00]

Well, I've started the second amp, but apparently I blew it on qty. On some of the caps I only ordered 2 instead of 4 (2 per board, 2 boards per amp). So while I'm ordering those I figured I might as well pick up some other caps to try in place of the input coupling cap and as ceramic replacements. Still sticking with digikey, so nothing too exotic, but hopefully moving up the food chain a bit.

So here's what I'm ordering to try.

C1: Input coupling cap, 5uF / 25V
Kemet R82 4.7uF 50V - Polyester PET Metalized, radial 5mm pitch should drop right in. Have read some good things about these.
Panasonic ECW-FD 4.7uF 450V - Polypropylene Metalized, big will have to bend the leads in.
I also have some Siemens 3.3uF Mylar, and 6.8uF Vishay Tantalex I could try.

C2: 150pF / 100V
Kemet PFR Polypropylene foil 150pF 250V
I also have some 330pF Siemens PP

C4: 20pF / 500V
Haven't come up with anything, I have some silver mica and interesting mil-spec ceramics

C6/17/21 0.001uF / 100V
I have a bunch of Siemens Polystyrene 1000pF / 630V, probably others to try.

C7: 47pF / 100V
Same as C4

C 8/9/20: 0.47uF / 100V
These are "tropical fish" mylar? Will leave in.

C12/15/16/19: 0.01uF / 100V
Kemet R46 Polypropylene Metalized 10,000pF 630V, should drop in w/10mm lead spacing
I also have some Panasonic ECQ-M (polyester foil?), and res caps to try

C13/14: 0.1uF / 100V
Panasonic ECW-FD Polypropylene Metalized 0.1uF / 450V
Kemet R71 Polypropylene Metalized 0.1uF / 420V
I also have some WonderCaps, and Panasonic ECQ-Ms to try

C18: 0.0033uF / 100V
Kemet R41 Polypropylene Metalized 3300pF 1000V
Panasonic ECW-H(A) Polypropylene Metalized
I also have Siemens 2700pF PP and 3900pF PS, silver mica 6800pF, and Panasonic ECQ-Ms to try

C22: 560p / 500V
Parallel Kemet PHE450 330pF 400V and Kemet PFR Polypropylene foil 220pF 400V
Not a lot of options, spec'd 10% so 470 and 680 are outside. I do have a few silver mica and those mil spec ceramics as well.

I don't know how much chance I'll actually get to experiment with any of these, but its nice to have options

 

[Moving Ahead]

Good question on Q1/3. Since thats's the input diff pair I assume that's pretty key. After a fair amount of cursing and knuckle bashing I was able to sneak some micro grabbers in and measure the Vbe on each channel. Measured with no input, here's what the pairs look like:

581 / 572 mV : 9mV dc offset

579 / 587 mV : 8mV dc offset

Its more of a gain thing. Matched gain at this point is the go. theres probably nothing wrong with them, (i thought i'd give you a worm.) if there were any transistors needing a check it'd be them. besides The dc trim-pot will compensate .

Agreed on the ceramics. I don't know how much difference it will make but its probably worth doing

For me i'm never surprised when I like the sound of an amp and I open it up and there's an absence of ceramic.

I like the idea of replacing the pots too... They do look a bit crudy,

Once you do this you'll never look back.

Any thoughts on what a reasonably priced improvement to electrolytic might be for the 5uF input coupling cap?

one less e-cap in the chain has to be a good thing. If you do replace don't get the over sized ones. there are nice stacked film available.

Another option is the use of COG ceramic if you want the original look. cheaper too. But you look like a guy that doesn't do things in halves.

 

[miscrms00]

Digikey arrived, so back in business. First board of second amp done, and de-ceramic'd.

This one uses pretty much all the Kemets from the list above. I cheated the parallel 220pF on C22 in on the back side to keep the top clean. The exceptions are the Siemens PS (C6/17/21) and military type ceramics for C7 and C4. I don't know for sure if these are any better than regular ceramics, they are all from a surplus lot I got years ago when my school EE department closed down their parts stores. Some of the bigger ones are marked C0G, some have other markings in other systems, but these small ones have no markings other than capacitance.

I figure I'll do the second board with the Panasonics. Then I'll have two boards with electrolytic input cap and original ceramics, and one each with the Kemet Polyster input and Kemet PP, and one with Panasonic PP. I'll probably try silver mica for the small value caps on the panasonic board if I can find the right values in my stash.

I haven't found a source yet for a pot that will fit into the pad locations (3/8" outer to outer), is sealed, and would be operable from all 4 directions. With the two rotations of the pot on each board, and one board mirrored relative to the other, you end up having to operate one from every orientation (top, bottom, left and right). Most pots these days seem to be on 0.1" pitch instead of 3/16" from what I've found on digikey, mouser, etc. Planning to go down to my local surplus place and see if I can come up with something. Anyone have any suggestions on these?

I also popped Q1 and Q3 off this board while I was waiting. They are both marked TZ-81. All I have currently to test is the Hfe measurement on my HF DVM, and the diode check function. Here's how the two devices measure:
Q1: Hfe = 283, Vbe = 0.714, Vbc = 0.711
Q3: Hfe = 265, Vbe = 0.709, Vcc = 0.707

 

[miscrms00]

Second board for the second amp is done!

This one uses the Panasonic ECWFD metalized polypropylene dips for the input couple and some of the higher value ceramic replacements. The low value ceramics are all replaced with NOS CDE silver mica dips. The 10nF and 1000pf are same as the other board, Kemet R41 metalized polypropylene and Siemens B31063 Polystyrene metal foil films.

Using the two different channels is maybe not ideal, but certainly a lot easier. It will be interesting to see if I can hear any difference between them, and between this amp the other with just the electrolytics redone including the input coupling cap.

 

[Moving Ahead]

Sorry if I seem a bit slow, but lets see if i got this straight.
You've got two stereo power amps.
Each power amp has two driver boards- one for the left and one for the right channel.
One of the amps you have used kemet film on both left and right driver boards.
The other amp you have used panasonic film on both left and right driver boards. (input cap seems big)
Am I right so far?
Can i ask what outcome/setup you are trying to achieve with the two (now different ) amps?

I also popped Q1 and Q3 off this board while I was waiting. They are both marked TZ-81. All I have currently to test is the Hfe measurement on my HF DVM, and the diode check function. Here's how the two devices measure:
Q1: Hfe = 283, Vbe = 0.714, Vbc = 0.711
Q3: Hfe = 265, Vbe = 0.709, Vcc = 0.707

That looks within spec(ish), but am i right theres another two on the other channel so there are 4 pieces that need matching? TZ-81 is low noise small signal type.
apart from that you look like a man on a mission. well done.

 

[miscrms00]

Sorry if I seem a bit slow, but lets see if i got this straight.
You've got two stereo power amps.
Each power amp has two driver boards- one for the left and one for the right channel.
One of the amps you have used kemet film on both left and right driver boards.
The other amp you have used panasonic film on both left and right driver boards. (input cap seems big)
Am I right so far?
Can i ask what outcome/setup you are trying to achieve with the two (now different ) amps?

Thanks, I really appreciate the feedback you've given so far. I have a rough sense of how things should work as an EE, but little practical audio experience. So the insights and questions bourn of experience are most welcome.

You're close-ish, here's a summary to clarify. The Altec 9440A is a pro 2 channel amp from the 70s that some claim was quite well done and very amenable to hi-fi use as is. Each channel has an identical driver board, so there are two per amp. I had one for a while before and liked it, but only replaced the main PS caps and didn't get to play with it long or try it in my system. Its not stereo, but not truly dual mono as I believe the +/- 80V power rails are shared. In my end use the two channels could be used as a L/R pair for one section, two sections of a center channel, or bridged to mono for a single channel section. Still up in the air. I do appreciate the heritage and history of these amps and don't want to muck with them too much, but also can see how their may be some changes due to age and different intended use that might be beneficial. I'm definitely not immune to getting sucked into the optimization vortex either

As far as what I have done so far, it doesn't make sense for final use. Just trying to come up with an expedient way to test differences between different configurations and see if there is a compelling reason to deviate from my original plan of replacing only electrolytics for reliability more than any expected improvement.

Amp 1: Both channels replaced all old electrolytics with new electrolytics, including input coupling cap. Original ceramics are still in place.

Amp 2 Ch A: Replaced all electrolytics as in Amp1, additionally changed out all ceramics. Most are Kemet metalized polypropylene film, 1000pf are Siemens polystyrene I had on hand. Input coupling cap is changed from electrolytic to Kemet stacked metalized polyester (R82) as I couldn't find a ~5uF polypropylene that would be a direct drop in.

Amp 2 Ch B: Replaced all electrolytics as in Amp1, additionally changed out all ceramics. Most are Panasonic metalized polypropylene film dips, 1000pF are same Siemens PS, 10nF is same Kemet PP as I didn't find a Panasonic in that size. All of the smaller value (<1000pF) ceramics are NOS CDE silver mica dips. The input cap is also Panasonic metalized polypropylene film dip, even though the size is a little ridiculous.

So, basically I have 4 identical mono driver boards with 3 different cap configurations that can be used in either position in either amp for testing in the workshop. This is a kind of lousy design-of-experiment as there are multiple variables changing on each board. But my thinking is if I can't hear any appreciable difference between the ceramics on the first two boards and the significantly different configurations on each of the second set of boards, then I don't really need to worry about it and can just pick one configuration and make all 4 boards the same. If there is a noticeable difference, and it seems promising, then I can worry about trying some more experiments to identify specific combinations to optimize. The only flaw I see in this plan is if you really need matched pairs on two channels to hear subtle differences in presence / sound stage, etc. Though honestly the system these are going into is so wacky and work in progress that I'm not sure that level of subtle improvement will be noticeable at this phase

That looks within spec(ish), but am i right theres another two on the other channel so there are 4 pieces that need matching? TZ-81 is low noise small signal type.

Correct. I only checked one input pair. There are three other input pairs on the other boards that have not been checked, as they have to be desoldered to check on my meter. I'm not too worried about it but thought it was worth checking a sample pair while I was waiting for more caps to arrive. I may well get paranoid and check them all at some point...

 

[miscrms00]

So, I had a little time tonight to do an initial comparison. I do most of my tinkering while the family is asleep, so quiet listening time is hard to come by

First there are lots of things that are not ideal about the comparison.
- Different caps on different driver boards, so transistors etc will have some differences too. Thought they are all the same age and seem original.
- Run on different amp channels, so there are other physical difference.
- Room setup is awful as I'm testing in the workshop full of junk.
- Pre-amp is less than ideal, two 40 y/o Altec 1608A mono mixers / amps with electrolytics replaced. Ok, but kind of noisy. Have to run hi and low boost and bright to sound decent.
- Speakers are decent, JBL S38, but not in the same league as the eventual application (vintage JBL Pro). One is a bit beat up but still seems fully functional.
- Lots of caps changed all at once between different boards.
- My youngest (2) came out to "help" and moved the mic between boards, I moved it back but position may not be identical.
- I don't really know what I'm doing

So with that all in mind, here's what I did. The nice thing is as I hoped there is enough slack in the connections that I can run the amp with the driver boards sitting on top for relatively quick change outs.

First I ran a pink noise FFT (iPhone AudioTools w/UMIK-1) of each channel first with the electrolytic only board and then boards with the electrolytics plus ceramic replacement and improved input couple caps.

Channel 1, green = electrolytic only change, mustard = electrolytics plus mostly panasonic PP and silver mica

Channel 2, green = electrolytic only change, mustard = electrolytics plus mostly Kemet PP and polyester input couple.

I don't see any real significant differences, particularly at the hi or low end, I don't see any (and didn't hear any) frequency extension. There are some possible differences in the midrange, but these could also be due to the mic move or measurement accuracy, or just that there's already a lot going on there in this setup.

Next I played a test track to see how they sounded. I picked Brubeck Take Five because I know it well enough to know how it should sound, and it has a lot of separation of instruments between the two channels so I could try to focus on the sound of each instrument through each channel as well as overall impression. I ran several cases. First with both boards with electrolytic only replacements. Then with the kemet board swapped into one channel. And then with the panasonic board swapped into the other channel. Then I swapped the R/L inputs to hear the other instruments through each board.

Results? Well, first I will say I didn't hear any thing super dramatic. But I will say when I installed the Kemet board I heard improvement, and then again when I installed the Panasonic board with it. I didn't hear much of any difference in the highs and lows, ride cymbal, bass drum, etc, but in the middle it did seem like there was more separation between instruments. Not so much spatially, that was pretty good in both cases, but sonically. IE one instrument didn't seem to mask another as much, which I suspect can be related to more or less distortion. In particular I've noticed the piano chords in the background can get kind of lost in some cases, and they definitely seemed more present when that channel was run with one of the ceramic and input coupling caps replaced. Not louder per se, just more distinguishable. The occasional string buzzes on the acoustic bass are another one I've noticed can get buried and were improved. As far as the two upgraded boards, I couldn't really hear much difference between the two at this stage. I also didn't hear anything worse about them than the electrolytic only boards.

So my plan going forward is to go ahead and do the input cap changes and ceramic swap outs on the other two boards in the same way I did the second set. That will give me a pair of each configuration to listen to and see if there is anything more I can distinguish between the two sets of caps.

 

[Moving Ahead]

I think I get the general idea of what youre doing. A real EE.
My biggest fear for you- is losing an output with your procedure. There looks to be a gazillion of them on the rear and that would hurt the ego to start replacing them.
A suggestion is to narrow down your final setup. They dont seem very domestic friendly, but I'm sure you can figure out a way to convince the wife.

These are power amps, with the changes youre doing you might squeeze the last 2% out of them at best. If you want night and day shades you need to look at the earlier stages. The power amp section is doing what its designed to do, and yours looks like a complicated circuit.

Looks like there are differences in the mid range with the film changes. Which is your critical listening. As for kemet or panasonic, you'll need a set of golden ears, and proper listening environment. Personally i think you'll be splitting hairs.

In the end I think it'd be best to have two identical boards in the unit, Mr Altec would have done it that way.

The bumblebee caps above the VU meters do drift.

You might get the little one to move the mike out 200 paces for testing and open 'em right up into their sweet spot.

If your looking for a night and day comparison you could try ZTX694B (center Base) for the diff input. They are not technically low noise but they might just open up or they might not...........to be or not to be.
TZ-81
http://www.datasheetarchive.com/dl/5c723f9765a319e26d471db77caca31e764afb/O/TZ-81

Don't forget to take the little tyke outside for an ice cream.

 

[miscrms00]

Finished modifying the boards from the first amp giving me a pair of each configuration. You can see in the bottom right board I was out of the 150pF CDE silver mica and had to use a larger (physically) SRC silver mica.

Each pair of boards was put into amp 1, and adjusted for bias and output null. I hadn't been able to adjust bias previously, as I haven't had a clamp meter capable of measuring DC. Based on reviews I ended up ordering a GTC CM-100 through Amazon. Not cheap, but still inexpensive compared to a Fluke or the like. After zeroing out the measurement I have to say I'm pretty impressed with its ability to measure 1mA DC resolution. Seems very repeatable and fairly insensitive to position.

After tuning up each board I ran a 2 channel FFT sweep for each pair. Same as before, AudioTools on iPhone with a Umik-1 using pink noise, averaged about 30s with 1/3 octave smoothing applied.

There may be some slight gain difference between the two pairs, but leaving that aside there does seem to be some more clear differences. Again trying to ignore the periodic structure that I think is room related, the Panasonic / Mica board set does seem to have several dB higher output in the mid range, and at high frequency including some high frequency extension. Listening to both sets again with Brubeck, some clear differences were also audible, again mostly in the midrange. First of all both sets sounded very good overall. But I did notice with the Kemet board the sax had a noticeable harshness which seemed much smoother on the Panasonic boards. I am assuming the main difference is the Polyester Kemet vs. the Polypropylene Panasonic used in the input coupling cap, but it would take more experimentation to know for sure.

In the interest of getting these amps into my system and moving on to other projects I will probably mod the Kemet boards to Panasonic and call it good. If I get inspired I might try changing just the input couple cap and retesting....

 

[miscrms00]

Agreed Moving Ahead, I hadn't really been looking to optimize but got intrigued. My main concern is reliability. I do think there were some worthwhile improvement, probably due mostly to changing the input coupling cap from electrolytic to film as discussed above. So its probably time call it a win and move on I've still got some clean up work to do on my McInstosh MA-230 before I can finally reshuffle all the amps in the downstairs rack. Then I still have the new JBL 2404H tweeters and servo-drive subwoofer to play with integrating. Lots still to do, but I figure I can always circle back on the amps

Thanks again for the suggestions!

 

[Binkman]

i must admit very nice board work and you can modify off circuits possibly for aux inputs to other amp, too. even stack a board if room in the box. Nice photos all around.
bink

 

[Moving Ahead]

Good to see you've got the results you're after.

 

[miscrms00]

Just a brief followup for completeness. I swapped the input coupling caps on the "kemet" board from the R82s to the Panasonics and had a quick listen. I couldn't hear much of any difference between the two sets of boards, so I do think that points to the input cap being the dominant effect. It also occurred to me that the 3 mylar tropical fish caps are probably those that Altec thought were the most significant with regard to sound/tone, and I haven't touched those. I don't think I will be doing that any time soon, but its probably something to think experimenting with in the future.