• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

A Tube Amp Restoration Walk Through

Welcome aboard! We're glad to have you as a partner.

And, if this thread is any example of the fine work you do, we're really going to look forward to more posts.

Especially of your work on Fisher tube gear! :yes:
 
tcdriver said:
Thank you so much for posting your cleaning and restoration process. The detailed description of your process and pictures are very helpful. I do have a question about the power supply capacitors you added. It looks like, as best I can tell from your picture, that you added capacitance in parallel to the existing capacitors. Is that correct or did you disconnect the originals? If you did parallel the new and old, did you check the original capacitors for leakage? Thanks again for the excellent write-up.
Click to expand...

The capacitors we added were not put in parallel to the originals. What we do is to cut the connections to the original old-fashioned can capacitors so we can leave them in place to preserve the original aesthetics of the vintage amps. We do not recommend leaving any of the old electrolytic capacitors in vintage gear.

Even if they test good, they are certainly nearing their end of life and should be replaced. As you probably know, well-made modern caps are much smaller, have lower ESR, higher temperature tolerance and longer life expectancy than the vintage technology. So we take advantage of this by mounting them under the chassis while leaving the disconnected cans in place to preserve the original vintage look. (We also recommend polishing the cans to make them look new again)
 
As there has been much debate about submersion cleaning ruining electronic components, we thought we would post all the performance graphs we made for this Fisher X-100B after it was ultrasonically cleaned and restored/modified. As is evident from the graphs, this amplifier not only is in perfectly good working condition, it now performs better than factory specifications. We tried to include a short description of each graph and how to read it, but please let us now if you have any questions, and I apologize for the small size of each graph:

Screen-Shot-2015-05-06-at-9.24.30-AM.png


X100B1.png


X100B2.png


X100B3.png
 
AEA Audio said:
The capacitors we added were not put in parallel to the originals. What we do is to cut the connections to the original old-fashioned can capacitors so we can leave them in place to preserve the original aesthetics of the vintage amps. We do not recommend leaving any of the old electrolytic capacitors in vintage gear.

Even if they test good, they are certainly nearing their end of life and should be replaced. As you probably know, well-made modern caps are much smaller, have lower ESR, higher temperature tolerance and longer life expectancy than the vintage technology. So we take advantage of this by mounting them under the chassis while leaving the disconnected cans in place to preserve the original vintage look. (We also recommend polishing the cans to make them look new again)
Click to expand...
Got it. Thank you.
 
I do same with old cans myself. cut em loose and place new repolacements undr chassis.
Makse for orig.look and no blown up caps later.
AEA does nice work.
 
AEA --I thoroughly applaud your efforts to provide measured performance results for the units you have restored, as that is the only scientific way to represent that a given unit has in fact been restored to factory performance specifications. It is something largely missing from this industry, but something I've done for decades, so I applaud when others have the same standards as well!

If I may however, may I suggest that a tweaking of your graphs may provide more useful information? In particular:

1. The graph on page 2 is for THD versus power, but yet no indication of what frequency that graph is run at is given. This is almost surely results performed at 1 kHz -- which is entirely in keeping with the rather loose way Fisher typically specified this performance criteria in their specifications offered. I would suggest that the test frequency should be identified, as distortion varies greatly with frequency as no doubt you know.

2. The graph on page 3 is of frequency response at max power. Frequency response tests are historically performed at a 1 watt power level (certainly by Fisher back in the day), so as to remove the power bandwidth element from the equation. Running the test at maximum power then brings this element into play. While the Fisher model that is the subject of the graph most likely has a flat response down to 20 Hz at a 1 watt level, any response at that frequency at an implied maximum power output of 26 watts RMS would be completely unusable due to the gross distortion levels produced -- and of course, it could never produce anywhere close to this power level at 20 Hz anyway. I would suggest then that this test would provide a more accurate representation by running and specifying the power level of the test at 1 watt, or specifying specifically what "max power" represents in this graph.

3. Finally, on page 5, the graph has THD measured against frequency, but now, power level is missing. Now granted, Fisher did not provide this information for the X-100B, but again, the graph would carry more weight if the power level was known, as it would look quite different at a power level of say within 1 db of rated power output -- which is how such curves were traditionally published. I would suggest then that the power level also be included that the graph was run at.

Again, I offer all of this only as thoughts to improve an already excellent service you perform. It may be that newer performance renderings conform to the way you have presented them. But conducting your tests in a manner that is consistent with the way they were performed back in the day would provide the clearest picture of the results from your restoration services, versus the specification provided by the manufacturer.

Thanks for entertaining these thoughts, best wishes for your company, and welcome to AK!

Dave
 
Hiya,

dcgillespie said:
AEA --I thoroughly applaud your efforts to provide measured performance results for the units you have restored, as that is the only scientific way to represent that a given unit has in fact been restored to factory performance specifications. It is something largely missing from this industry, but something I've done for decades, so I applaud when others have the same standards as well!

If I may however, may I suggest that a tweaking of your graphs may provide more useful information? In particular:

1. The graph on page 2 is for THD versus power, but yet no indication of what frequency that graph is run at is given. This is almost surely results performed at 1 kHz -- which is entirely in keeping with the rather loose way Fisher typically specified this performance criteria in their specifications offered. I would suggest that the test frequency should be identified, as distortion varies greatly with frequency as no doubt you know.

2. The graph on page 3 is of frequency response at max power. Frequency response tests are historically performed at a 1 watt power level (certainly by Fisher back in the day), so as to remove the power bandwidth element from the equation. Running the test at maximum power then brings this element into play. While the Fisher model that is the subject of the graph most likely has a flat response down to 20 Hz at a 1 watt level, any response at that frequency at an implied maximum power output of 26 watts RMS would be completely unusable due to the gross distortion levels produced -- and of course, it could never produce anywhere close to this power level at 20 Hz anyway. I would suggest then that this test would provide a more accurate representation by running and specifying the power level of the test at 1 watt, or specifying specifically what "max power" represents in this graph.

3. Finally, on page 5, the graph has THD measured against frequency, but now, power level is missing. Now granted, Fisher did not provide this information for the X-100B, but again, the graph would carry more weight if the power level was known, as it would look quite different at a power level of say within 1 db of rated power output -- which is how such curves were traditionally published. I would suggest then that the power level also be included that the graph was run at.

Again, I offer all of this only as thoughts to improve an already excellent service you perform. It may be that newer performance renderings conform to the way you have presented them. But conducting your tests in a manner that is consistent with the way they were performed back in the day would provide the clearest picture of the results from your restoration services, versus the specification provided by the manufacturer.

Thanks for entertaining these thoughts, best wishes for your company, and welcome to AK!

Dave
Click to expand...

Dave this is why you are truly a first class asset to AK.

Frannie
 
I aplaud measured performance records .
I cut my audio teethe reading all the perf. data from all the mfrs and clinics from all the mags. we could see what we we were getting.
But today the garage tek dont got the bucks for all the necessary gear to do it right as its quite spensiv IMO. I got a scope, funct gen, dmm, ac rms vm, loads and most of the basic stuff for testing and fixing but certainly not freq sweeps and the like or spectrum analysis as AEA does.
Should have bought all the gear I would need when I worked at HP as it was all right there in front of me and employee discounts were common. But all I got was a scope.
 
Don't be so hard on yourself Ferninando. A good tech who is well versed in using their scope can do a very good job in troubleshooting, and adjusting for optimum performance -- and you're a very good tech!

Dave
 
dcgillespie said:
AEA --I thoroughly applaud your efforts to provide measured performance results for the units you have restored, as that is the only scientific way to represent that a given unit has in fact been restored to factory performance specifications. It is something largely missing from this industry, but something I've done for decades, so I applaud when others have the same standards as well!

If I may however, may I suggest that a tweaking of your graphs may provide more useful information? In particular:

1. The graph on page 2 is for THD versus power, but yet no indication of what frequency that graph is run at is given. This is almost surely results performed at 1 kHz -- which is entirely in keeping with the rather loose way Fisher typically specified this performance criteria in their specifications offered. I would suggest that the test frequency should be identified, as distortion varies greatly with frequency as no doubt you know.

2. The graph on page 3 is of frequency response at max power. Frequency response tests are historically performed at a 1 watt power level (certainly by Fisher back in the day), so as to remove the power bandwidth element from the equation. Running the test at maximum power then brings this element into play. While the Fisher model that is the subject of the graph most likely has a flat response down to 20 Hz at a 1 watt level, any response at that frequency at an implied maximum power output of 26 watts RMS would be completely unusable due to the gross distortion levels produced -- and of course, it could never produce anywhere close to this power level at 20 Hz anyway. I would suggest then that this test would provide a more accurate representation by running and specifying the power level of the test at 1 watt, or specifying specifically what "max power" represents in this graph.

3. Finally, on page 5, the graph has THD measured against frequency, but now, power level is missing. Now granted, Fisher did not provide this information for the X-100B, but again, the graph would carry more weight if the power level was known, as it would look quite different at a power level of say within 1 db of rated power output -- which is how such curves were traditionally published. I would suggest then that the power level also be included that the graph was run at.

Again, I offer all of this only as thoughts to improve an already excellent service you perform. It may be that newer performance renderings conform to the way you have presented them. But conducting your tests in a manner that is consistent with the way they were performed back in the day would provide the clearest picture of the results from your restoration services, versus the specification provided by the manufacturer.

Thanks for entertaining these thoughts, best wishes for your company, and welcome to AK!

Dave
Click to expand...

1) Yes, you’re right. Thanks for pointing that out, we overlooked it. It should be specified in the report. It is at 1kHz, the normal specification. Don’t know how we overlooked it, but we will make that correction right away.

2) You are right again. The manufacturers back then often intentionally measured frequency response at low power to avoid power bandwidth issues. However, because this was misleading, we use a power bandwidth measurement technique for the graph. The FTC now requires power bandwidth to be specified 3 db below max rated power. In our current reports, we specify this more informative 3 db below max rated power for that graph.

3) Right, we mistakenly left out the power, which is 3 db below max rated power: Same as used in the frequency response graph.

Dave, thanks very much for taking the time to help us improve the reports. It is very much appreciated!
 
I thought I would share some more photo results of ultrasonic cleaning on a Pilot 602 receiver that I just finished.

I've organized it so that the before ultrasonic cleaning photos are on the left, and the post ultrasonic cleaning photos on the right:


IMG_6010.jpg
IMG_6018.jpg


IMG_6011-e1431545635432.jpg
IMG_6022-e1431545551403.jpg


IMG_6013.jpg
IMG_6023.jpg


IMG_6012.jpg
IMG_6024.jpg


IMG_6021.jpg

I hope you can see how ultrasonic cleaning is not only about removing tarnish and gunk from the chassis, but is also, and more importantly, about removing tarnish from components.
 
Hey, you left a spot!!!!!!!!


:D :D :D :D

Seriously, only see one pic. And while I was an early skeptic, I'm coming around to becoming a believer. Maybe even do some business in the future.

Welcome aboard!
 
This is an eyeopening thread for me (as it is for many others) about cleaning the vintage amps. Maybe we should do a group buy on large enough ultrasonic cleaner. If not, I'd be happy with purchasing a smaller one just for those darn nuts and bolts with painful tarnish and oxidation.
 
You must log in or register to reply here.
Back
Top Bottom