Hi All
Like many others on here i have some armstrong audio love to rekindle, i remember as a child my uncle playing his 625 receiver through some kef speakers, at that time it was so lovely to hear and look at, this was my first introduction to hifi.
Even today the looks of these amplifiers are brilliant! So i went and searched and bought these two units, the 625 and 621.
I want to rebuild the integrated 621 amplifier, the 625 was thrown in for the deal as it was damaged in transit. For me this was great as i can use it for spares, they share the same makeup right?
So this is my plan, i would love to upgrade the boards with better components such as doing a full recap of all the boards such as the preamp, amplifier board and psu etc
I also want to change any transistors and diodes which i have read can go faulty. I would like to build this amplifier to be like brand new and to last another 20-30 years.
So my questions are:
Is it worth replacing resistors? What resistors were used (type carbon?Metal film etc) The amp board would be easy to do i could replace all the components with new parts (not to sure about the bias resistors) they look odd and special heat disspersation types.
The amp board has other tiny driver transistors, do these tend to go and would it be a good idea to replace these?
I am slowly going through the circuit diagrams to understand it all better, i also know over the many years they changed components and always upgraded or made changes.
I will start with the 621 by just replacing all the capacitors, diodes and small transistors to get it working and running as it should. The 625 will be where i play around with it and try to upgrade the components.
Any advice would be appreciated such as what the bias should be, where to measure the bias, known things that should be checked, changed, replaced etc.
One other thing is the potentiometers on the 621 are very stiff unlike the 625 which feel smooth to turn, the input selector on the 621 takes quite a bit of force to click into place, where as the 625 is smooth and doesn't take me to squeeze my fingers hard to turn it. Wondering if spraying detox in there will solve this issue.
From what i have read the 600 range were modular and share the same boards
C28 input board
A17 power board
Z20 power supply
which means i could actually swap boards from the 625 to the 621? The 625 seems to be more modern.
Also can anyone tell me good replacement parts for these parts
BC267 & BC297
BC384L transistors (would like similar or better i.e lower noise)
The bias trimmers seem old and risky to use would like to replace these with some bourns what power value would suffice here they come as 1/5w to 0.5w.
On the A17 power board there are some axial capacitors
2 x 220uf 50v and 2 x 47uf 100v which are polarised, could i replace these with non polarised (i have some which are of good quality and make).
The z20 power board has some tiny 1n4003 rectifying diodes could i replace these with something better like schottky diodes?
I also see many small eirie type capacitors of small values, thinking these would need replacing aswell? think these are Poly film types. Could i replace these with polyester, polystyrene or silver micas? lots in the tone control board!
The psu smoothing cap is 3300uf would it be ok to go slightly bigger? say 4000-6000uf? same goes for the DC speaker output caps which are 4000uf wondering if going slightly bigger has an advantage?
For further information!
The Armstrong 600 Series - Amplifier FAQ and diagrams
Yes its worth spending time and money to rebuild this back to its former glory! Any tips or things to look out for would be great.
Thanks all those interested!
Like many others on here i have some armstrong audio love to rekindle, i remember as a child my uncle playing his 625 receiver through some kef speakers, at that time it was so lovely to hear and look at, this was my first introduction to hifi.
Even today the looks of these amplifiers are brilliant! So i went and searched and bought these two units, the 625 and 621.
I want to rebuild the integrated 621 amplifier, the 625 was thrown in for the deal as it was damaged in transit. For me this was great as i can use it for spares, they share the same makeup right?
So this is my plan, i would love to upgrade the boards with better components such as doing a full recap of all the boards such as the preamp, amplifier board and psu etc
I also want to change any transistors and diodes which i have read can go faulty. I would like to build this amplifier to be like brand new and to last another 20-30 years.
So my questions are:
Is it worth replacing resistors? What resistors were used (type carbon?Metal film etc) The amp board would be easy to do i could replace all the components with new parts (not to sure about the bias resistors) they look odd and special heat disspersation types.
The amp board has other tiny driver transistors, do these tend to go and would it be a good idea to replace these?
I am slowly going through the circuit diagrams to understand it all better, i also know over the many years they changed components and always upgraded or made changes.
I will start with the 621 by just replacing all the capacitors, diodes and small transistors to get it working and running as it should. The 625 will be where i play around with it and try to upgrade the components.
Any advice would be appreciated such as what the bias should be, where to measure the bias, known things that should be checked, changed, replaced etc.
One other thing is the potentiometers on the 621 are very stiff unlike the 625 which feel smooth to turn, the input selector on the 621 takes quite a bit of force to click into place, where as the 625 is smooth and doesn't take me to squeeze my fingers hard to turn it. Wondering if spraying detox in there will solve this issue.
From what i have read the 600 range were modular and share the same boards
C28 input board
A17 power board
Z20 power supply
which means i could actually swap boards from the 625 to the 621? The 625 seems to be more modern.
Also can anyone tell me good replacement parts for these parts
BC267 & BC297
BC384L transistors (would like similar or better i.e lower noise)
The bias trimmers seem old and risky to use would like to replace these with some bourns what power value would suffice here they come as 1/5w to 0.5w.
On the A17 power board there are some axial capacitors
2 x 220uf 50v and 2 x 47uf 100v which are polarised, could i replace these with non polarised (i have some which are of good quality and make).
The z20 power board has some tiny 1n4003 rectifying diodes could i replace these with something better like schottky diodes?
I also see many small eirie type capacitors of small values, thinking these would need replacing aswell? think these are Poly film types. Could i replace these with polyester, polystyrene or silver micas? lots in the tone control board!
The psu smoothing cap is 3300uf would it be ok to go slightly bigger? say 4000-6000uf? same goes for the DC speaker output caps which are 4000uf wondering if going slightly bigger has an advantage?
For further information!
The Armstrong 600 Series - Amplifier FAQ and diagrams
Yes its worth spending time and money to rebuild this back to its former glory! Any tips or things to look out for would be great.
Thanks all those interested!
Last edited:
I have refurbished a few of these and found that the ITT capacitors are still as good as new.
Replace the 3000u main smoothers and the loudspeaker coupling caps with up to 4k7u and no more! Aluminium Capacitors | RS Components
The power supply will struggle if you go too high in value.
After replacing the main capacitors, run them to check for performance. They should be on a par with the B&O amps of that time.
Check the preset current skeleton pots for tarnish and clean them if needed.
Set up the quiescent current and they should go on for another 50 years.
Replace the 3000u main smoothers and the loudspeaker coupling caps with up to 4k7u and no more! Aluminium Capacitors | RS Components
The power supply will struggle if you go too high in value.
After replacing the main capacitors, run them to check for performance. They should be on a par with the B&O amps of that time.
Check the preset current skeleton pots for tarnish and clean them if needed.
Set up the quiescent current and they should go on for another 50 years.
Thanks Jon, thats why i am covering all angles and replacing components that do tend to fail, i especially use turntables and i know transistors of age can get noisey and with capacitors drifting.
If i can cover myself and replace known parts that fail this indeed will last another 50years!
While its all apart i would like to upgrade parts.
If i can cover myself and replace known parts that fail this indeed will last another 50years!
While its all apart i would like to upgrade parts.
There's an argument for updating the resistors that affect linearity and noise on the main amps, basically the feedback network and input network - metal film replacements for those if you want.
Thanks Mark, my thoughts were that resistors have come along way since the 70s, something like prp or vishay Dale RN are superior low noise metal film types, thoughts were this would be ideal.
Checking through the resistors with my DMM suprisingly most of the resistors are in spec!
Checking through the resistors with my DMM suprisingly most of the resistors are in spec!
Congratulations on finding a classic not burnt to toast already or repaired to uselessness.
I'm running the US copy (done badly) the dynaco ST120, up to 14 hours a day. At least the surviving parts. These had inadequate heat sinks and fried outputs frequently until I added more fins, and fans.
I wouldn't monkey with small transistors except for cause, like frying noises or popping noises. These noises were common in 1966, but if you got ones that don't do it leave them. Taking parts out & replacing can damage lands on boards, one channel of mine burnt so frequently that I've had to replace the board with point to point.
Plastic film caps post 1964 and disc caps don't deteriorate either.
Electrolytic caps do age badly. You could buy a peak ESR meter and change for cause, but the damage is so universal I change them all in any amp that has parts over 20 years old. If you go bigger on the rail caps you need to put a NTC thermistor in the transformer primary to keep from damaging things like the switch with a huge turn on surge. Tantalum caps age badly, also. Buy 10000 hours or at least 3000 hour service life caps, you don't want to do this again. Distributors are full of 500 hour and 1000 hour caps for repair shops that want the customer to come back with the same problem in two years. Panasonic, Rubicon, Nichicon, Vishay, Kemet, Aerovox have been premium product, but they DO sell short life caps for the high demand for them.
Resistors, there has been a technology upgrade since the sixties. Carbon comp resistors 100 kohms & over have audible hiss, thermal noise. I've achieved obvious hiss improvements by going to metal film. I don't monkey with the small ones unless they look burnt or are off value. Warning, metal film has been downsized recently to run hot enough to burn boards. Don't use any metal film smaller than 1 watt. I get good resistors from vishay, welwyn, multicomp(farnell), dale.
Pots are unreliable due to oxidation of the wipers. New ones aren't better than old ones, IMHO. Clamp the bias control pot with a parallel diode, schottky if the voltage developed is under .4. I can't quote resistor number, I had to throw my armstrong 62x schematics away because a virus infected all my gifs. Volume and tone control pots are also obvious wear items, but don't replace if they are not causing problems.
The 1n4003 on the power board make less turn off noise than schottky diodes, believe it or not. Leave them alone. Ultra fast also cause problems. The edge is sharper and has higher frequency components. If the design doesn't have a .01 uf 1000 v disk cap across the bridge output, install one. If it does have one, these have often been stressed by lightning surges and could also stand to be replaced. Warning 1000 v disk caps are hard to find these days. Old TV's are my best source.
Transistors, if you do have a problem I find any On semi part with the right pinout and a Vceo at least double the in circuit voltage, to be extremely quiet and quite high gain. Sometimes I have to degenerate the gain with a collector to base resistor, the gains these days are so high. Watch changing TO18 or TO5 with TO92, the wattage may be excessive. I tend to add heat sink to my drivers even though they didn't need it to survive an FTC amp power test, as I leave my amps on 14 hours a day frequently. The armstrong had nice external heat sinks for the output transistors, IMHO.
Best of luck.
I'm running the US copy (done badly) the dynaco ST120, up to 14 hours a day. At least the surviving parts. These had inadequate heat sinks and fried outputs frequently until I added more fins, and fans.
I wouldn't monkey with small transistors except for cause, like frying noises or popping noises. These noises were common in 1966, but if you got ones that don't do it leave them. Taking parts out & replacing can damage lands on boards, one channel of mine burnt so frequently that I've had to replace the board with point to point.
Plastic film caps post 1964 and disc caps don't deteriorate either.
Electrolytic caps do age badly. You could buy a peak ESR meter and change for cause, but the damage is so universal I change them all in any amp that has parts over 20 years old. If you go bigger on the rail caps you need to put a NTC thermistor in the transformer primary to keep from damaging things like the switch with a huge turn on surge. Tantalum caps age badly, also. Buy 10000 hours or at least 3000 hour service life caps, you don't want to do this again. Distributors are full of 500 hour and 1000 hour caps for repair shops that want the customer to come back with the same problem in two years. Panasonic, Rubicon, Nichicon, Vishay, Kemet, Aerovox have been premium product, but they DO sell short life caps for the high demand for them.
Resistors, there has been a technology upgrade since the sixties. Carbon comp resistors 100 kohms & over have audible hiss, thermal noise. I've achieved obvious hiss improvements by going to metal film. I don't monkey with the small ones unless they look burnt or are off value. Warning, metal film has been downsized recently to run hot enough to burn boards. Don't use any metal film smaller than 1 watt. I get good resistors from vishay, welwyn, multicomp(farnell), dale.
Pots are unreliable due to oxidation of the wipers. New ones aren't better than old ones, IMHO. Clamp the bias control pot with a parallel diode, schottky if the voltage developed is under .4. I can't quote resistor number, I had to throw my armstrong 62x schematics away because a virus infected all my gifs. Volume and tone control pots are also obvious wear items, but don't replace if they are not causing problems.
The 1n4003 on the power board make less turn off noise than schottky diodes, believe it or not. Leave them alone. Ultra fast also cause problems. The edge is sharper and has higher frequency components. If the design doesn't have a .01 uf 1000 v disk cap across the bridge output, install one. If it does have one, these have often been stressed by lightning surges and could also stand to be replaced. Warning 1000 v disk caps are hard to find these days. Old TV's are my best source.
Transistors, if you do have a problem I find any On semi part with the right pinout and a Vceo at least double the in circuit voltage, to be extremely quiet and quite high gain. Sometimes I have to degenerate the gain with a collector to base resistor, the gains these days are so high. Watch changing TO18 or TO5 with TO92, the wattage may be excessive. I tend to add heat sink to my drivers even though they didn't need it to survive an FTC amp power test, as I leave my amps on 14 hours a day frequently. The armstrong had nice external heat sinks for the output transistors, IMHO.
Best of luck.
Last edited:
I just really want to get this singing again, and if i can make some minor improvements from component choices to replacing known faulty parts then happy to do so!
from my own experiences those little outhouse transistors after a few decades do get noisey, so would rather replace them.
Also there are these diodes which are ITT44 any good suitable replacements for them?
from my own experiences those little outhouse transistors after a few decades do get noisey, so would rather replace them.
Also there are these diodes which are ITT44 any good suitable replacements for them?
I'd check for noise with a sound probe before changing transistors. For one thing, looks as if you have TO5 transistors with heat sinks around them. These are EBC and you can't buy them anymore except counterfeits from *****. The wattage is not 500 mw like TO92, it is up to 10 W with a heat sink on them, which you seem to have a lot of. These midpower transistors have been replaced by TO126 which has a flat back, and is BCE. Also TO220 which has a flat back and is BCE. One of my TO220 replacements that I twisted around to fit for a TO5 broke the leg off after a half dozen years in service. Only thing close I've found in stock recently (farnell) is the 2n5682 which is TO18 EBC and not quite as much power to the heat sink as a 1n2102.
In sixties to eighties organs that have hundreds of diodes in the keying networks, maybe one a year around here needs replacement. So I wouldn't monkey with the IT44. Not unless you found the same DC voltage on both sides of one.
Diodes that can age badly are zener diodes. Also voltage regulators based on internal zener diodes. Look in the power supply areas for those. All the germanium zeners in my schober recital organ (designed 1962) were bad. The builder gave me a mausoleum of dead zeners in little original boxes he bought from the shober co - an obvious design flaw if ever there was one. I put in a silicon 3 watt zener. The Armstrong should have silicon diodes.
100 ma generic diodes are 1n4148 (rice grain sized, formerly 1n914), half to one amp ones are 1n4003 to 1n4007. 6 amp ones come from diodes inc under a house number, see RS or farnell selector tables. On semi sells some 3 amp diodes, that have a 8 mm diameter case. 6 amp diodes are about 10 mm diameter case. Don't buy fast, ultra fast, or schottky diodes, the plain jane ones produce less RF interference from turning off because the RF components of the fourrier transform of the vertical edge are lower in frequency. Where you do want speed is VAS & driver transistors. 30 mhz Ft ones (MJE15028/29) sound better than 3 mhz ones (TIP41c/42c). Cymbals & top octave piano sound better with faster drivers & vas. The original RCA TO5 drivers & vas in my st120 (one board survived the burnups) are quite good sounding on high freq sources. The TIP41/42c on the point to point board slightly less so.
In sixties to eighties organs that have hundreds of diodes in the keying networks, maybe one a year around here needs replacement. So I wouldn't monkey with the IT44. Not unless you found the same DC voltage on both sides of one.
Diodes that can age badly are zener diodes. Also voltage regulators based on internal zener diodes. Look in the power supply areas for those. All the germanium zeners in my schober recital organ (designed 1962) were bad. The builder gave me a mausoleum of dead zeners in little original boxes he bought from the shober co - an obvious design flaw if ever there was one. I put in a silicon 3 watt zener. The Armstrong should have silicon diodes.
100 ma generic diodes are 1n4148 (rice grain sized, formerly 1n914), half to one amp ones are 1n4003 to 1n4007. 6 amp ones come from diodes inc under a house number, see RS or farnell selector tables. On semi sells some 3 amp diodes, that have a 8 mm diameter case. 6 amp diodes are about 10 mm diameter case. Don't buy fast, ultra fast, or schottky diodes, the plain jane ones produce less RF interference from turning off because the RF components of the fourrier transform of the vertical edge are lower in frequency. Where you do want speed is VAS & driver transistors. 30 mhz Ft ones (MJE15028/29) sound better than 3 mhz ones (TIP41c/42c). Cymbals & top octave piano sound better with faster drivers & vas. The original RCA TO5 drivers & vas in my st120 (one board survived the burnups) are quite good sounding on high freq sources. The TIP41/42c on the point to point board slightly less so.
Last edited:
Indianjo you have an absolute mountain of good advice, which has given me homework to figure out what you have written, still struggling to compare specification of transistors, what type, the values and what would be a suitable replacement without it causing any imbalance in the circuits. Yes I would like to enhance and better what Armstrong have done as I believe it has the capability to do so.
Difference is I would like to be able to change all the diodes, transistors to new parts. This enables me to gather information for other future Armstrong 600 enthusiasts for replacement parts that work.
At the moment having to read other threads and forums for recommendations on like for like parts.
The output transistors are functional but would like to know or figure out direct replacements for these aswell.
Starting to desolder the capacitors and choose suitable replacements.
Thanks
Difference is I would like to be able to change all the diodes, transistors to new parts. This enables me to gather information for other future Armstrong 600 enthusiasts for replacement parts that work.
At the moment having to read other threads and forums for recommendations on like for like parts.
The output transistors are functional but would like to know or figure out direct replacements for these aswell.
Starting to desolder the capacitors and choose suitable replacements.
Thanks
Hi!
My advice is to make changes in different steps.
First change the electrolytic caps (absolutely necessary !) You should hear in difference.
Next step, you can change the PS diodes for some with higher power.
You can then change the low power transistors in the preamp.
I wouldn't change the output transistors, if they do their work fine.
Transistors, as resistors, film ou ceramic caps, are almost eternal.
These components age and degrade over time, but there's no rule.
So IMHO the best advice is to replace the faulty or noisy components only, and leave the others.
You have more chance (?) to get counterfeit new components to replace the good old ones ...
My advice is to make changes in different steps.
First change the electrolytic caps (absolutely necessary !) You should hear in difference.
Next step, you can change the PS diodes for some with higher power.
You can then change the low power transistors in the preamp.
I wouldn't change the output transistors, if they do their work fine.
Transistors, as resistors, film ou ceramic caps, are almost eternal.
These components age and degrade over time, but there's no rule.
So IMHO the best advice is to replace the faulty or noisy components only, and leave the others.
You have more chance (?) to get counterfeit new components to replace the good old ones ...
Actually there is a rule about resistors. Carbon comp resistors over 100k can get moisture incursion over the years, changing the value. Exception are resistors made on US mil spec lines with water resistant paint made after about 1961. These resistors are not those. For example US built Hammond organs resistors are pretty good, the ones built in the Netherlands look just like them and wander off value anywhere near the sea.
Output transistors, depends on the build date. Original RCA select 2n3055 transistors have a 400 khz Ft and limit high frequency accuracy. Simulations have shown dynaco ST120's with TIP3055 or later output transistors do much better at high frequency intermodulation distortion tests than the originals. Nearly anything made after 1975, say MJ15015, will do great in this respect. I'm using NTE 181 copies of the MJ15003, (to go by the soa spec) from about 1989 and they sound great on top octave piano solos cymbals & bells. Warning when using high Ft output transistors, you have to install 1 ohm 1 watt (or greater wattage) resistors between the driver and the OT base to prevent oscillation.
On the small signal transistors, one problem with the "direct replacement" transistor companies like NTE, central semi, new jersey semi, is that the most of the datasheets leave out all the specifications. You don't even know if you are getting an amplifier transistor or a switching transistor, that doesn't show up in most specs, just as a layout choice of the artwork. Has to do with the linearity of the gain at different collector currents. In terms of replacement, I've found On semi MPS06/56 and MJE15028/29 can replace most any small signal low power amplifier transistor, but the pinouts are WRONG for EBC parts and the TO92 MPS06/56 have to be carefully evaluated that the power and current limits are not exceeded since TO5 was so powerful. The old TO5 heat sinks won't fit, either.
As said by alcha if the 1n4003 are limiting the power, modern 6 amp replacements are available in standard turn off grade like the Diodes inc A6. Usually home use peak power may be 70 w/ch but average power is about two watts. The heat sinks won't allow a lot more power without adding a fan. I found this out by trying to play a church with a keyboard and the ST120. After 3.5 hour rehearsal the base resistor to the output transistor went up in a ball of flame under the lecturn. A real 100w amp for church use, the Allen S100, has the same two output transistors as the ST120, less rail voltage, and a heat sink 10000 times bigger.
I second the comment about using counterfeit new components to replace good old ones. There are hundreds of 1960's small signal transistors in these old organs I play, 99.9% of them are fine. Vox and shober organ germanium transistors excepted.
Last edited:
Yes Alcha that is what i am planning to do, Ive ordered all the electrolytics for all boards, ill replace them all then go about testing each section and get it running and working, i am talking about the armstrong 621 here.
At present i have stripped the armstrong 625 down.
These are the boards comprising of the amplifier board, input board and control/tone control board.
these are the boards i want to mess about with and replace the components with new types. There is so much potential here, the ceramic disc capacitors annoy me, dont ask me why i just feel they are cheap parts, id rather use something better like silver mica or Polyester. Even the electrolytics can be replaced with elna silimics ii or nichicon audio specific parts.
I would even like to change all those little erie capacitors on the tone/control board!
This is the 621 amplifier board you will notice the different transistors and capacitor values.
The large ITT capacitors which are 50v 470uf & 100v 47uf can i use bipolar capacitors here? Do they have to be electrolytic?
This is the z20 power board from the 625 look at the rectifying diodes, this are much bigger then the rice grain size ones on the 621!
I intend to test that these are in spec and swap them over the 621, or if you can suggest suitable new replacements i would rather buy new.
I would ultimately like to upgrade all the 625 boards with better newer components, after all these amplifiers and armstrongs philosophy was to always improve these amplifiers "subject to continuous development" so i figure i would do the same, hoping you guys can help me develop it a little further!
So to start with what would be people choices be for these replacements:
A17 Board transistors
BC267 & BC297
ITT44 & 1N4148
BC297&BC267
BF258
2N5322 & 2N5320
40636
These are taken from the circuit diagram, i am slowly myself figuring out what would be a suitable exchange for these components, i would like to be able to just by them from farnell, mouser etc.
You can still source a number of original parts it seems:
LittleDiode – Electronic Components for All
Need to find something similar to:
BC384L its been suggested that perhaps the BC547 or the lower-noise BC549 may suffice.
Last edited:
Armstrong 621 rebuild - Members Pictures - Speakers, Electronics, etc. - The Classic Speaker Pages Discussion Forums
this was a nice read
this gentleman replaced the original PL4003 diode bridge with 1N5408 devices and original 3300UF/100V PSU cap was replaced with 6800UF unit.
this was a nice read
this gentleman replaced the original PL4003 diode bridge with 1N5408 devices and original 3300UF/100V PSU cap was replaced with 6800UF unit.
Your 625 rectifiers appear to be 8 mm in diameter, or probably 3 amp ones.
If you really want to experiment with new transistors and diodes, I suggest you make or buy new boards and put in the old chassis. That way you won't damage the old lands and the old unmatchable transistors diodes. This is relaxed enough layout to build point to point on Nema C material (garolite McMaster-Carr), if you don't want to learn to operate layout software (I don't). I use kynar or teflon insulated wire to avoid burnup with an unwatched soldering iron. It took me two schematics to find one that worked as well as the not-burned up original, which needs 7 auxillary transistors on a separate board to sound good. Armstrong 621 did not have the cold lower idle bias current (output transistors) problem of the ST120, circuit shouldn't need changing.
I've started cutting leads off old axial lead caps, bending hooks, and soldering radial lead ones to them, instead of risking damaging the board by pulling the leads out. The caps can be held down with glue.
We bought a swell engine board to replace a 1984? one that caught fire (connector pin back poked through transformer lead insulation & shorted power ciruit). The new board they sent us, $180, decent price, doesn't work. It is trying to go forwards & backward, fast and slow, all at the same time, instead of picking one mode. 3 motion LED's of 4 on at once. It has 28 transistors the same number as the old ones. They leak like sieves under actual voltage 24v. I've replaced 23 transistors so far and damaged two board lands that needed repaired with a wire patch. New transistors are BD139/140 from Farnell - 80 v to92 with matching ECB pinout . They are multicomp, which is farnell's house brand, but I've found their house QA to be adequate in matching the spec numbers they publish.
If you really want to experiment with new transistors and diodes, I suggest you make or buy new boards and put in the old chassis. That way you won't damage the old lands and the old unmatchable transistors diodes. This is relaxed enough layout to build point to point on Nema C material (garolite McMaster-Carr), if you don't want to learn to operate layout software (I don't). I use kynar or teflon insulated wire to avoid burnup with an unwatched soldering iron. It took me two schematics to find one that worked as well as the not-burned up original, which needs 7 auxillary transistors on a separate board to sound good. Armstrong 621 did not have the cold lower idle bias current (output transistors) problem of the ST120, circuit shouldn't need changing.
I've started cutting leads off old axial lead caps, bending hooks, and soldering radial lead ones to them, instead of risking damaging the board by pulling the leads out. The caps can be held down with glue.
We bought a swell engine board to replace a 1984? one that caught fire (connector pin back poked through transformer lead insulation & shorted power ciruit). The new board they sent us, $180, decent price, doesn't work. It is trying to go forwards & backward, fast and slow, all at the same time, instead of picking one mode. 3 motion LED's of 4 on at once. It has 28 transistors the same number as the old ones. They leak like sieves under actual voltage 24v. I've replaced 23 transistors so far and damaged two board lands that needed repaired with a wire patch. New transistors are BD139/140 from Farnell - 80 v to92 with matching ECB pinout . They are multicomp, which is farnell's house brand, but I've found their house QA to be adequate in matching the spec numbers they publish.
Last edited:
With the wires all cut you'll have a hard time testing the boards after every 2 parts changed to see if you made the performance better or worse. That is what I do. Focuses your eyes on exactly where you installed a problem, testing after every 2 parts.
Look & see if you can turn the board upside down on a book for parts installation, then back right side up for testing. Peavey's work that way.
I did put a connector on wires to a hammond chassis instead of those turret solder joints wrapped with wire so I could remove & replace quickly over & over. TE connectivity Mate-n-lock commercial (250 v). Comes in up to 15 pins. Crimping those pins with pliers is a PIT* but I'm not buying the $250 official tool. DB25 connectors are rated for about 40 volts, beware using them on rail voltage.
The ceramic caps short out RF and ultrasonic oscillation better than anything else. Silver mica is close if you have to change them. But ceramics have an initial quality problem that should have been sorted out by Armstrong incoming inspection. They don't deteriorate unless subject to lightning surges on the line, IMHO. The ones on the schematic appear to be anti-oscillation parts, nowhere near the AC input.
I wouldn't monkey with the "erie" caps which appear to be film.
Audio grade electrolytics are IMHO a big scam. The prices are rediculous. I buy industrial grade brands I listed above in the longest service life I can find at an authorized distributor. Havent received a dud yet. Some I installed in 2009 after I quit working still work fine, in amps & radios I use 14 hours a day.
The 47 & 100 uf caps have to be electrolytic because of size. Parts-express sells film caps in those sizes for speaker crossovers, but they are HUGE and about $15 apiece. Probably wouldn't fit in that case.
The bipolar cap you want to use, will probably work. I have never seen a bipolar cap for sale that has a service life quoted. I think they must all be sealed with red gum. Go ahead if you want to replace this again in 3-5 years. 10000 hour electroltytic caps have some serious sealant used to keep the water from evaporating out.
Little diode is not an authorized parts distributor like farnell, RS, mouser, digikey, some german firms I can't remember. Any reputation they have comes from testimony on the net, which has been counterfeited before. I don't have the incoming inspection equipment to test transistors for bogus parts, except the Vceo leakage current test which the transistors from the swell engine vendor failed. Waste of time IMHO.
This whole project has the smell of a boat anchor. Good luck finishing.
Do you even have a sound probe to test for noisy transistors? One is an amplifier & speaker combo, protected on the front end by back to back red LED's series a 1000 ohm resistor each from input to signal ground. This clamps the input voltage to 1.4 max.Ahead of those series with the input and signal ground are .047 600 v film caps, to keep DC out of your amplifier & clamp. Building one of those & learning how to use it is more important than changing all your transistors & diodes for no cause but they are old.
If you want reference level sound build 2 channels of a honey badger. Sold up top in the diyaudio store on the banner. This circuit will never go under .2 to maybe .1% harmonic distortion. Fine for my $600 apiece SP2 speakers, but if you're going to hear .002% distortion you'll need magnaplaners or better and ears that still respond to 20000 hz. pretty rare in a man, my ears only go to 14000 hz. New transistors are not going to tweak the distortion out of a circuit that has no constant current sources, active bias control, high gain VAS, or centering servo.
Last edited:
These are my thoughts indianjo.
As i said i have an Armstrong 625 and 621.
The 621 i am just replacing the electrolytics, thats every single one which will be a mixture of Elna silmic II and Vishay.
I also want to change the trimmers for the bias, dont want to risk them not working and causing issues. How many mw would suffice?
The 625 is merely for learning and practicing, its a hobby for me to learn as i am still learning. I like vintage hifi but everytime i buy anything of age components always need changing! I had a leak stereo 20, quad ii's, yamaha ca 600, krell ksa's etc. Every bloody one of them needed servicing. I've also built a few class a amplifiers very simple ones like jean hiraga 8w and 30w from kits.
Im no expert, i just like building, taking things apart, trying new things and ideas. I learn as i go along, but also help from various forums is a big help. I work full time and when i have spare time i do audio stuff.
I like Armstrong 600 series amplifiers, i like the old skool ethos. I should be able in my head replace every component like for like in terms of value and specifications for each and every component.
Im not changing the circuit at all, just want to see if it would work and if it would perhaps make it perform better. Along the way i get to learn more which is a bonus, and probably afterwards i will know the armstrong 600 series amps like the back of my hand.
As i said i have an Armstrong 625 and 621.
The 621 i am just replacing the electrolytics, thats every single one which will be a mixture of Elna silmic II and Vishay.
I also want to change the trimmers for the bias, dont want to risk them not working and causing issues. How many mw would suffice?
The 625 is merely for learning and practicing, its a hobby for me to learn as i am still learning. I like vintage hifi but everytime i buy anything of age components always need changing! I had a leak stereo 20, quad ii's, yamaha ca 600, krell ksa's etc. Every bloody one of them needed servicing. I've also built a few class a amplifiers very simple ones like jean hiraga 8w and 30w from kits.
Im no expert, i just like building, taking things apart, trying new things and ideas. I learn as i go along, but also help from various forums is a big help. I work full time and when i have spare time i do audio stuff.
I like Armstrong 600 series amplifiers, i like the old skool ethos. I should be able in my head replace every component like for like in terms of value and specifications for each and every component.
Im not changing the circuit at all, just want to see if it would work and if it would perhaps make it perform better. Along the way i get to learn more which is a bonus, and probably afterwards i will know the armstrong 600 series amps like the back of my hand.
The current across the bias current trimpot if BF258 was shorted would be ~1.8 ma. I^R=P=.032 W so a standard 125 mw pot should work.
As I said previous, no trimpot is safe. Clamp the trimpot (parallel) a diode if the voltage is above .4 and use a schottky diode parallel if the voltage you end up with is under .4. That way the idle current can't run away any more than that.
The quality problems with 1966 transistors tended to be popcorn noise and frying bacon noise from process problems, particularly contaminants. You can hear those with the sound probe. If you don't have those noises, you don't have a problem requiring transistor replacement, imho. Or diode either.
The problem with modern transistors is that the pinouts are all wrong the are not EBC. They are BCE or EBC. I've found twisting the legs around to make fit leads to broken legs after a few years. Don't do it if you don't have to. Also modern transistors don't come in metal cans and are touchy about too much power or voltage. The heat sinks you have DONT FIT the new transistors. So you have to calculate a lot of power dissipations, which is not easy.
I agree that aged electrolytic caps are trash. I agree that carbon comp resistors over 100k are noisy. I agree that the cheap carbon comp resistors this armstrong has are suspect anywhere near the sea, like North London on a tidal estuary. I agree volume pots and sometimes tone control pots can be wearout problems. Switches and connectors can get oxidized and block sound. Other than that I'd investigate with the sound probe to hear further problems before fixing them.
Best of luck.
As I said previous, no trimpot is safe. Clamp the trimpot (parallel) a diode if the voltage is above .4 and use a schottky diode parallel if the voltage you end up with is under .4. That way the idle current can't run away any more than that.
The quality problems with 1966 transistors tended to be popcorn noise and frying bacon noise from process problems, particularly contaminants. You can hear those with the sound probe. If you don't have those noises, you don't have a problem requiring transistor replacement, imho. Or diode either.
The problem with modern transistors is that the pinouts are all wrong the are not EBC. They are BCE or EBC. I've found twisting the legs around to make fit leads to broken legs after a few years. Don't do it if you don't have to. Also modern transistors don't come in metal cans and are touchy about too much power or voltage. The heat sinks you have DONT FIT the new transistors. So you have to calculate a lot of power dissipations, which is not easy.
I agree that aged electrolytic caps are trash. I agree that carbon comp resistors over 100k are noisy. I agree that the cheap carbon comp resistors this armstrong has are suspect anywhere near the sea, like North London on a tidal estuary. I agree volume pots and sometimes tone control pots can be wearout problems. Switches and connectors can get oxidized and block sound. Other than that I'd investigate with the sound probe to hear further problems before fixing them.
Best of luck.
Last edited:
Hi Guys,
I know this thread is over 2 years old but you seem to have expeirence of the Armstong Amps, I have an Armstrong 525, my fathers old amp I am trying to resurrect, it had a transistor issue that kept blowing one of the 1A fuses on the power board, this has now been resolved with swapping out some transistors from an old ebay fined, now the right channel has continuos thumping/thudding in the speaker, no audio. Checked the voltage on the speaker terminal and it had a reading of 25vdc other chachannel 0.14vdc. Did some google'ing and thought it could be the capcitors related to this channel, de-soldered all caps and all read roughly the same, multimeter reads 1.566-1.466 mf (caps say 1250uf 30vdc).
Any ideas suggestions, not to technical!!!
Thanks
I know this thread is over 2 years old but you seem to have expeirence of the Armstong Amps, I have an Armstrong 525, my fathers old amp I am trying to resurrect, it had a transistor issue that kept blowing one of the 1A fuses on the power board, this has now been resolved with swapping out some transistors from an old ebay fined, now the right channel has continuos thumping/thudding in the speaker, no audio. Checked the voltage on the speaker terminal and it had a reading of 25vdc other chachannel 0.14vdc. Did some google'ing and thought it could be the capcitors related to this channel, de-soldered all caps and all read roughly the same, multimeter reads 1.566-1.466 mf (caps say 1250uf 30vdc).
Any ideas suggestions, not to technical!!!
Thanks
25 vdc on speaker indicates the speaker cap is shorted out. It may not read shorted with a DVM ohms scale or capacitance scale; the test voltage is 2 v or less. But under the half of 82 v the speaker cap runs, it can be shorting. If you can't afford the $5 for a new 3900 100 v cap, put a 4.7 k resistor series the milliamp reading of your DVM, parallel combination to the 82 v. If the milliamps don't read near zero, you have a short.
Besides, electrolytic caps that read about 25% high in value have been dried up in my experience. A $120 ESR meter would prove it, but I find an amp's box of replacement capacitors to be a lot cheaper than that. If you don't want to do the job again in 10000 operating hours, buy e-caps with service life >3000 hours. Farnell & digikey will let you see the service life in the selector table if you ask for it. I don't know about RS, they don't operate here. I re-e-capped a ST70 4 times in 22 years with average life store shelf e-caps. Too much work.
Replacing 40636 output transistor with modern output transistors can cause ultrasonic oscillation. You won't hear it but it could damage your speakers.
Old 40636 had 400 khz Ft, modern ones have 2 to 4 mhz Ft. Seriously you must check the result with a scope after that change, or look with an ANALOG AC voltmeter blocked to ground with a 390 pf cap. If you read any AC voltage through a 390 pf cap, it is ultrasonic. DVM can't see radio frequencies.
There are some alleged 2n3055H transistors for sale in the US by surplus house electronicsurplus.com . I bought 4. But I haven't checked suitability for 82 v rail yet and I can't check Ft at all. 40636 were selected by RCA for 41 v Ice leakage suitability from the 2n3055 production stream. RCA fab is now a county park in New Jersey.
Notice Mr. Vishalk never reported the success of his "upgrade". I suspect replacing "nearly everything" produced a garden ornament, not a functioning amplifier.
Besides, electrolytic caps that read about 25% high in value have been dried up in my experience. A $120 ESR meter would prove it, but I find an amp's box of replacement capacitors to be a lot cheaper than that. If you don't want to do the job again in 10000 operating hours, buy e-caps with service life >3000 hours. Farnell & digikey will let you see the service life in the selector table if you ask for it. I don't know about RS, they don't operate here. I re-e-capped a ST70 4 times in 22 years with average life store shelf e-caps. Too much work.
Replacing 40636 output transistor with modern output transistors can cause ultrasonic oscillation. You won't hear it but it could damage your speakers.
Old 40636 had 400 khz Ft, modern ones have 2 to 4 mhz Ft. Seriously you must check the result with a scope after that change, or look with an ANALOG AC voltmeter blocked to ground with a 390 pf cap. If you read any AC voltage through a 390 pf cap, it is ultrasonic. DVM can't see radio frequencies.
There are some alleged 2n3055H transistors for sale in the US by surplus house electronicsurplus.com . I bought 4. But I haven't checked suitability for 82 v rail yet and I can't check Ft at all. 40636 were selected by RCA for 41 v Ice leakage suitability from the 2n3055 production stream. RCA fab is now a county park in New Jersey.
Notice Mr. Vishalk never reported the success of his "upgrade". I suspect replacing "nearly everything" produced a garden ornament, not a functioning amplifier.
Last edited: