Well, after about 20 years sitting in the attic I pulled my old Dynaco Pat 4 out but had forgot I gave the Stereo 35 away to a friend years ago. I put the Pat 4 together when I was in Junior High School way back in the sixties but I was never able to afford a Stereo 120. I picked one up the other day ago off eBay for eighty bucks. It looks goo and plays Hendrix OK but there is some hum coming from my old Klipsch Model H's with this amp. I love this old Dynaco stuff so I think I'll rebuild it. Where to start. I figure the best place would be to replace the old Electrolytic capacitors. The output caps - C7s - are 3300 mfd / 50 volts. There 2 inches in diameter. The three power capacitors, C9, C11,& C12, are 100 mfd/100 v; 500 mfd/500v; and 3300mfd/100v, respectfully. Where can I get these cans in this day and age? Or can I just replace them with new smaller caps? Who are the preferable manufactures of Capacitors now? Shurely, capacitor design, materials and quality has changed. Thanks for all your help.
I used modern >3000 hour service life caps, which are much smaller. I had to glue them down with silicon seal instead of matching the old clamps. I use newark.com and mouser.com, 470 uf instead of 500, the only iffy brands they sell are CDE and Epcos. Nichicon, Rubicon, Panasonic, all sell great "snap in" caps as long as you don't buy the 1000 hour cheapos sealed with gum rubber. United Chemicon have been okay but they come from a country known for lieing a lot on QC paperwork. The others come from coutries that make the news less for imitation beef, imitation baby formula, imitation concrete, etc. I like to purchase long service life e-caps; my ST70 needed new caps every 4-6 years for low power and low B+ voltage, back when you bought parts from the front of the distributor because the salesmen didn't have time to talk to anybody that didn't have a 50 part order. Now we have the internet, you can actually see the specs of what you are buying.
There are also some caps on the PC14 board that need replacing. the 5 uf input caps I replace with plastic film. See the pwgtang circuit before buying all those NP and small electrolytics on the board, though.
The ST120 has a little too much crossover distortion at low volume cold, the circuit was a very early design. See pwgtang's improvement (documented in this thread a couple of years ago) which swaps about 10 parts to make the circuit more conventional and faster to "warm up" to good sound. He deletes the capacitor coupling from input to driver stage, deletes the 330 ohm resistors, and makes a conventional double diode + potentiometer arrangement between the bases of the driver transistors. I put the extra parts on a cinch 8 terminal strip above the PC14's so I don't have to cut the boards up too much. I also bring the top and bottom of the .5 ohm resistor between the output transistors up to the terminal strip so I can measure the bias current (about 20-40 ma is about right) to adjust it.
I've used my repaired unmodified ST120 in PA filling a 150 seat church for Christmas 1990, and got it hot enough to melt solder on the output capacitors. The solid core wire sprung up, hit the cover, and blew the output transistors. So those no fins "heat sinks" have to be supplemented. I originally put two PCAT power supply fans blowing on the two output flanges, bolted to a channel under the amp which replaces the broken feet, and powered by a 9 VDC wall transformer. I've since added a very flat multi finger heat sink to the middle of the output transistor flange.
Some people think the 2n3772 or 40xxx RCA transistors (selected RCA 2n3055 for higher voltage) are too slow to reproduce high frequencies (14 khz to 20 khz) since they have Ft of 200 khz instead of 4 mhz of modern transistors. I had to replace my TO3 transistors about 1988 with NTE equivalents of MJ15003, (NTE60) but I would use MJ15015 these days. I bought my amp all blown up and repaired it without the schematic . Don't replace the regulator board with a modern transistor, the low gain of the old one is key to getting shutdown at 6.75 amps delivered to the output transistors. I tested mine with a 8 ohm log resistor, to set the trip point since my first attempt with a NTE60 had it tripping at ~ 2 amps. If you use faster transistors, install the output resistor and the 50 pf base to emitter capacitors of the dynaco "TIP mod" documented in greg Dunn's dynaco history website. Faster transistors could oscillate, and I have done that, tacking the capacitors to the bottom of the board without drilling any holes.
You can buy a kit from djoffe with new PC boards that converts the ST120 to a LM3886 IC amp, (akita company) but I think his 3/8" solid no fingers "heat sinks" are worse than the old dynaco ones if you leave it turned on for more than x hours. I like my solution better, I leave mine on 24 hours at a time sometimes, and 18 hours every day, with the unit installed behind a bookcase to cover the noise of the fans.
I think the higher gain of the NTE60 transistors (or MJ115003 or later) allow the amp to deliver maybe 100 w/ch or greater with the 72 v power supply as long as you don't do it too long and overheat it. Say the cannon booms of 1812 overture, lots of boom there. My scope died after I came up with this theory and it wasn't a one shot memory scope anyway, so I have no proof. But my upgraded ST120 is as loud as my CD800s (360 W/ch @ 8 ohm) on 8 ohm Peavey ST2.XT speakers in my living room where base music volume setting is about 1.5 Vpp.
Have fun.
There are also some caps on the PC14 board that need replacing. the 5 uf input caps I replace with plastic film. See the pwgtang circuit before buying all those NP and small electrolytics on the board, though.
The ST120 has a little too much crossover distortion at low volume cold, the circuit was a very early design. See pwgtang's improvement (documented in this thread a couple of years ago) which swaps about 10 parts to make the circuit more conventional and faster to "warm up" to good sound. He deletes the capacitor coupling from input to driver stage, deletes the 330 ohm resistors, and makes a conventional double diode + potentiometer arrangement between the bases of the driver transistors. I put the extra parts on a cinch 8 terminal strip above the PC14's so I don't have to cut the boards up too much. I also bring the top and bottom of the .5 ohm resistor between the output transistors up to the terminal strip so I can measure the bias current (about 20-40 ma is about right) to adjust it.
I've used my repaired unmodified ST120 in PA filling a 150 seat church for Christmas 1990, and got it hot enough to melt solder on the output capacitors. The solid core wire sprung up, hit the cover, and blew the output transistors. So those no fins "heat sinks" have to be supplemented. I originally put two PCAT power supply fans blowing on the two output flanges, bolted to a channel under the amp which replaces the broken feet, and powered by a 9 VDC wall transformer. I've since added a very flat multi finger heat sink to the middle of the output transistor flange.
Some people think the 2n3772 or 40xxx RCA transistors (selected RCA 2n3055 for higher voltage) are too slow to reproduce high frequencies (14 khz to 20 khz) since they have Ft of 200 khz instead of 4 mhz of modern transistors. I had to replace my TO3 transistors about 1988 with NTE equivalents of MJ15003, (NTE60) but I would use MJ15015 these days. I bought my amp all blown up and repaired it without the schematic . Don't replace the regulator board with a modern transistor, the low gain of the old one is key to getting shutdown at 6.75 amps delivered to the output transistors. I tested mine with a 8 ohm log resistor, to set the trip point since my first attempt with a NTE60 had it tripping at ~ 2 amps. If you use faster transistors, install the output resistor and the 50 pf base to emitter capacitors of the dynaco "TIP mod" documented in greg Dunn's dynaco history website. Faster transistors could oscillate, and I have done that, tacking the capacitors to the bottom of the board without drilling any holes.
You can buy a kit from djoffe with new PC boards that converts the ST120 to a LM3886 IC amp, (akita company) but I think his 3/8" solid no fingers "heat sinks" are worse than the old dynaco ones if you leave it turned on for more than x hours. I like my solution better, I leave mine on 24 hours at a time sometimes, and 18 hours every day, with the unit installed behind a bookcase to cover the noise of the fans.
I think the higher gain of the NTE60 transistors (or MJ115003 or later) allow the amp to deliver maybe 100 w/ch or greater with the 72 v power supply as long as you don't do it too long and overheat it. Say the cannon booms of 1812 overture, lots of boom there. My scope died after I came up with this theory and it wasn't a one shot memory scope anyway, so I have no proof. But my upgraded ST120 is as loud as my CD800s (360 W/ch @ 8 ohm) on 8 ohm Peavey ST2.XT speakers in my living room where base music volume setting is about 1.5 Vpp.
Have fun.
Last edited:
Another ST120 nutter... good on you.
Owned mine since I bought it as a Kit in '71/2 ?
It gave years of Service Make no mistake though a Stock st120 is a truly crappy amp.. I tried various mods over the decades, with small joy.
Recently fitted the Updatemydynaco.com St120 refit Kit.
One reuses only the case and TX.
Note though the oem TX is potentially problematic, as it makes lotsa odd noises mechanical and electrical.
Upgrade Kit is easy and surprisingly inexpensive.
A true Bargain IMO.
Result was beyond expectations.. It's now a seriously Good Amp.
A Fine piece of design work by Mr Joffe.
It's main limitations are in the quality of the source(s) and the Speakers it powers.
It's good enough so that Garbage in = Garbage out.. seriously.
In fact it is now replacing a surprisingly expensive Pedigree dual amp setup. Surprised me.
NO heat issues whatsoever.. But then My speakers are 94db and the resultant sounds are exquisite.
Pat4 upgrades' were /are a mixed bag.
The Update upgrades Did seriously improve the things performance.. a lot.
However it wasn't one of Dyna's 'better' products.. possibly their very worst.
After too much time fooling with it's many foibles. It's now relegated to permanent storage use.
In hindsight my Pat 4 adventure was a complete waste of time.. Hopefully you might have better opinions.. or not 🙂
Good luck
Owned mine since I bought it as a Kit in '71/2 ?
It gave years of Service Make no mistake though a Stock st120 is a truly crappy amp.. I tried various mods over the decades, with small joy.
Recently fitted the Updatemydynaco.com St120 refit Kit.
One reuses only the case and TX.
Note though the oem TX is potentially problematic, as it makes lotsa odd noises mechanical and electrical.
Upgrade Kit is easy and surprisingly inexpensive.
A true Bargain IMO.
Result was beyond expectations.. It's now a seriously Good Amp.
A Fine piece of design work by Mr Joffe.
It's main limitations are in the quality of the source(s) and the Speakers it powers.
It's good enough so that Garbage in = Garbage out.. seriously.
In fact it is now replacing a surprisingly expensive Pedigree dual amp setup. Surprised me.
NO heat issues whatsoever.. But then My speakers are 94db and the resultant sounds are exquisite.
Pat4 upgrades' were /are a mixed bag.
The Update upgrades Did seriously improve the things performance.. a lot.
However it wasn't one of Dyna's 'better' products.. possibly their very worst.
After too much time fooling with it's many foibles. It's now relegated to permanent storage use.
In hindsight my Pat 4 adventure was a complete waste of time.. Hopefully you might have better opinions.. or not 🙂
Good luck
Last edited:
Another ST120 nutter... good on you.
Owned mine since I bought it as a Kit in '71/2 ?
It gave years of Service . I tried various mods .
Recently fitted the Updatemydynaco.com St120 refit Kit.
One reuses only the case and TX.
Easy and surprisingly inexpensive.
A true Bargain IMO.
Result was beyond expectations.. It's now a seriously Good Amp.
A Fine piece of design work by Mr Joffe.
It's main limitations are in the quality of the source(s) and the Speakers it powers.
It's good enough so that Garbage in = Garbage out.
In fact it is now replacing a a surprisingly expensive Pedigree dual amp setup.
NO heat issues whatsoever.. But then My speakers are 94db and the resultant sounds are exquisite.
Pat4 upgrades' were /are a mixed bag.
The Update upgrades Did seriously improve the things performance.. a lot.
However it wasn't one of Dyna's 'better' products.. possibly their very worst.
After too much time fooling with it's many foibles. It's now relegated to permanent storage use.
Good luck
Owned mine since I bought it as a Kit in '71/2 ?
It gave years of Service . I tried various mods .
Recently fitted the Updatemydynaco.com St120 refit Kit.
One reuses only the case and TX.
Easy and surprisingly inexpensive.
A true Bargain IMO.
Result was beyond expectations.. It's now a seriously Good Amp.
A Fine piece of design work by Mr Joffe.
It's main limitations are in the quality of the source(s) and the Speakers it powers.
It's good enough so that Garbage in = Garbage out.
In fact it is now replacing a a surprisingly expensive Pedigree dual amp setup.
NO heat issues whatsoever.. But then My speakers are 94db and the resultant sounds are exquisite.
Pat4 upgrades' were /are a mixed bag.
The Update upgrades Did seriously improve the things performance.. a lot.
However it wasn't one of Dyna's 'better' products.. possibly their very worst.
After too much time fooling with it's many foibles. It's now relegated to permanent storage use.
Good luck
Fun, Fun, Fun till your daddy takes the T-Bird Away .....
Oh I can see this Bulletin Board is going to be a lot of fun ....
I grew up in the basement playing with my dads Hammerlund Sp600 Receiver and a huge ham shop of Army surplus "stuff". My dad was one of the original eighteen year old "dit-dah" guys off a battle ship in WW2 and a life Army Mars / Ham . We had draws of resistor, chokes, relays, caps, silicon - boxes of tubes - 6L6s a plenty. I think most of that stuff went to others Hams and some of it to the dump. That got me into Dyna Equipment because I could put it together myself and the specs were terrific. I'm having a time today reading up on the Stereo 120 with the cover off, looking at whats been changed. (There's two big ceramic wire wound resistors bolted vertical , wired to each output post. There not on any diagrams I've found yet, Maybe a TIP? ) I guess I'm going to have to pull this amp apart after I get some caps.
On the other note: The boards Atikita is making sure look good. That looks like a great winter project to build a "120" from scratch or if I can find a good blown parts chassis with a good transformer. The prices are not bad either. Milling my own heat sinks out of a chunk of aluminum would be no problem for me - after I model it in Autocad.
I retired last year and now I don't have time to do things I don't want to do !
Oh I can see this Bulletin Board is going to be a lot of fun ....
I grew up in the basement playing with my dads Hammerlund Sp600 Receiver and a huge ham shop of Army surplus "stuff". My dad was one of the original eighteen year old "dit-dah" guys off a battle ship in WW2 and a life Army Mars / Ham . We had draws of resistor, chokes, relays, caps, silicon - boxes of tubes - 6L6s a plenty. I think most of that stuff went to others Hams and some of it to the dump. That got me into Dyna Equipment because I could put it together myself and the specs were terrific. I'm having a time today reading up on the Stereo 120 with the cover off, looking at whats been changed. (There's two big ceramic wire wound resistors bolted vertical , wired to each output post. There not on any diagrams I've found yet, Maybe a TIP? ) I guess I'm going to have to pull this amp apart after I get some caps.
On the other note: The boards Atikita is making sure look good. That looks like a great winter project to build a "120" from scratch or if I can find a good blown parts chassis with a good transformer. The prices are not bad either. Milling my own heat sinks out of a chunk of aluminum would be no problem for me - after I model it in Autocad.
I retired last year and now I don't have time to do things I don't want to do !
1000 ohms across the output jack is part of the "TIP mod" dynaco came up with when output transistors got faster. I used 3 W metal film resistors from multicomp. I believe the output resistor is to increase stability. the TI TIP3055 transistor was one of the first faster ones available, although it is not now generally available. MJ15003 and MJ15015 output transistors are still made in Mexico by ON Semi, purchaser of the Motorola plant. They have smaller screw holes, #4 instead of #6 for the 40xxx the amp came with in 1970, so you will need new mica washer, plastic ferrels, screws and nuts. I bought my insulator kit for "TO3" (not really) from newark.
My flat TO3 heatsink was a Thermaloy 506007B00000G. It fits the PC15 with the regulator transistor but doesn't fit the output transistors which are too low, but does bolt nicely out in the middle of the heat flange.
I have no trouble with the power transformer on my unit. In fact it took a lightning surge without damage that shorted the power switch on my PAS2 preamp. The transformer does not hum or rattle.
I've got the regulator board running at 6.7 amps trip point, but it runs open at 80 V rail instead of 72 per design, which is a heat problem covered by my fans. In fact, I had a djoffe feedback bias board installed that blew an input transistor and was running 240 ma idle current, which could have been a problem but wasn't with the fans. On to the pwgtang circuit. But with 80 v rails and high gain MJ15003 equivalent (NTE60) I could be getting 50% of the rail or 40 V peaks into my speaker, which could be 200 w. That is music power not RMS, due to the heat problem but I listen to a lot of classical music where a 70 db dynamic range is a worthy goal. Fire those cannon, Mr. Tschaikovsky! My 1812 Overture cannon doesn't clip IMHO even in the front yard on 4th of July.
Maintaining the output capacitor I believe to be a worthy goal, since I could have blown my speaker when the amp overheated and blew the output transistors, but it didn't. Blowing speakers due to DC is something the dynaco ST70 amp never would do, and I find that capability of 99.9% of all solid state amp rather shocking. I find the passion designers had to get rid of the output capacitor more of a cost cutting trick and a marketing tool, than improving sound. My ST120 with NTE60 transistors, the djoffe bias setting boards (not the LM3886 boards he sells), and the Peavey CS800s 260 w amp with .03% Harmonic Distortion below the clipping point, sound exactly the same at 1.5 Vpp output (base) on my best test record. My speakers are now 100 db Peavey SP2-XT's but when the amp overheated I was using Peavey 1210's. The CS800s is now tripping the breaker after 16 years hard bar band life; I find the ST120 a lot easier to work on. And making up the djoffe seven transistor bias boards and the pwgtang PC14 circuit modification gives me the feel of some do it yourself electronics. Buy four diodes, some cinch 8 terminal solder strip, two 100 ohm pots, and two red leds', your ready for the pwgtang circuit. The rest of the TIP mod from gregdunn's site requires four 47 pf disk capacitors.
I don't think there is room with the old wire coils for a full finned heat sink on each side, but with the smaller capacitors there is room to put one across where the old rail capacitor went, with 4 transistors on it. That way you could get away with one fan if you were using 20 W/ch base level in PA service.
I'm retired and listen to my amp and LP's CD's or FM radio while I'm not restoring old organs or playing them or the piano, in the winter or in the dark or rain anytime.
My flat TO3 heatsink was a Thermaloy 506007B00000G. It fits the PC15 with the regulator transistor but doesn't fit the output transistors which are too low, but does bolt nicely out in the middle of the heat flange.
I have no trouble with the power transformer on my unit. In fact it took a lightning surge without damage that shorted the power switch on my PAS2 preamp. The transformer does not hum or rattle.
I've got the regulator board running at 6.7 amps trip point, but it runs open at 80 V rail instead of 72 per design, which is a heat problem covered by my fans. In fact, I had a djoffe feedback bias board installed that blew an input transistor and was running 240 ma idle current, which could have been a problem but wasn't with the fans. On to the pwgtang circuit. But with 80 v rails and high gain MJ15003 equivalent (NTE60) I could be getting 50% of the rail or 40 V peaks into my speaker, which could be 200 w. That is music power not RMS, due to the heat problem but I listen to a lot of classical music where a 70 db dynamic range is a worthy goal. Fire those cannon, Mr. Tschaikovsky! My 1812 Overture cannon doesn't clip IMHO even in the front yard on 4th of July.
Maintaining the output capacitor I believe to be a worthy goal, since I could have blown my speaker when the amp overheated and blew the output transistors, but it didn't. Blowing speakers due to DC is something the dynaco ST70 amp never would do, and I find that capability of 99.9% of all solid state amp rather shocking. I find the passion designers had to get rid of the output capacitor more of a cost cutting trick and a marketing tool, than improving sound. My ST120 with NTE60 transistors, the djoffe bias setting boards (not the LM3886 boards he sells), and the Peavey CS800s 260 w amp with .03% Harmonic Distortion below the clipping point, sound exactly the same at 1.5 Vpp output (base) on my best test record. My speakers are now 100 db Peavey SP2-XT's but when the amp overheated I was using Peavey 1210's. The CS800s is now tripping the breaker after 16 years hard bar band life; I find the ST120 a lot easier to work on. And making up the djoffe seven transistor bias boards and the pwgtang PC14 circuit modification gives me the feel of some do it yourself electronics. Buy four diodes, some cinch 8 terminal solder strip, two 100 ohm pots, and two red leds', your ready for the pwgtang circuit. The rest of the TIP mod from gregdunn's site requires four 47 pf disk capacitors.
I don't think there is room with the old wire coils for a full finned heat sink on each side, but with the smaller capacitors there is room to put one across where the old rail capacitor went, with 4 transistors on it. That way you could get away with one fan if you were using 20 W/ch base level in PA service.
I'm retired and listen to my amp and LP's CD's or FM radio while I'm not restoring old organs or playing them or the piano, in the winter or in the dark or rain anytime.
Last edited:
Fun is when the work yields genuine result.
Having done many of the one time popular diy mods, plus some, to my long suffering 120.
With 20/20 hindsight/ honesty, the result was lipstick on a pig... period
Note that the 120 was/is one of (the?) most reviled amplifiers of all time.
An earned reputation.
Decision: do you lust after a never ending project to fill your time with
OR a very good sounding Amplifier to use.
Factoring in one's Audio expectations, quality threshold as well 😉
Having done many of the one time popular diy mods, plus some, to my long suffering 120.
With 20/20 hindsight/ honesty, the result was lipstick on a pig... period
Note that the 120 was/is one of (the?) most reviled amplifiers of all time.
An earned reputation.
Decision: do you lust after a never ending project to fill your time with
OR a very good sounding Amplifier to use.
Factoring in one's Audio expectations, quality threshold as well 😉
We can sit on our hands and let the lords of commerce sell us something new every 4 years - - - 3 years - - - 2 years - - - 6 months. The new Shimano shifter I bought in May is already destroyed, twisted out of line on a piece of grass in the gears. It looks like the old one that lasted 18 years but is that shiney stuff steel? Made in the same country that makes LM3886 IC's to look at the sticker on the baggie.
The parts count on the protection circuit of this PV-1.3k split supply amp that is so dangerous to speakers it came with a sticker "do not use channel A", is hitting 60, and that is just for detecting DC and stopping it. No parts for overcurrent detecting. The crowbar circuit on the PV-1.3k didn't work, it merely melted traces to the SCR on the board. The protection circuit in the Peavey CS800s looks good with 3 levels of protection, and the amp costs >$1000 new. I can't spend that every day. Listeners to my performances, don't pay me.
The emporer has no clothes. Amps without a speaker series cap are disasters waiting to happen. IMHO. My speakers cost $600 used, none of the amps was more than $100, the ST120 was $50 with incinerated OT's. The speaker cap in the ST120 did work when I overheated it in PA service for 150 people.
Do it yourself means I am in control, not some banker in NYC or Hong Kong. I got probably 2000 hours pleasant low distortion listening out of the modified ST120 before the djoffe feedback bias board failed. What really took it down, I kicked it and pulled some wires out, not internal failure. Maybe next time on the pwgtang circuit I'll get 10000 trouble free hours. The ST70 tube amp sure needs attention every 8000 hours, either caps, output tubes, or rectifier tube. And the vacuum tubes come mostly from Russia, our kind gentle partners in commerce, right? DIY, the parts to update my ST120 come mostly from Mexico, or oriental countries that are not building an island in their neighbor's sea to protect their oil grab.
You bare need to tighten the nuts on your transformer. Elastic stop nuts, maybe, time marches on.
The parts count on the protection circuit of this PV-1.3k split supply amp that is so dangerous to speakers it came with a sticker "do not use channel A", is hitting 60, and that is just for detecting DC and stopping it. No parts for overcurrent detecting. The crowbar circuit on the PV-1.3k didn't work, it merely melted traces to the SCR on the board. The protection circuit in the Peavey CS800s looks good with 3 levels of protection, and the amp costs >$1000 new. I can't spend that every day. Listeners to my performances, don't pay me.
The emporer has no clothes. Amps without a speaker series cap are disasters waiting to happen. IMHO. My speakers cost $600 used, none of the amps was more than $100, the ST120 was $50 with incinerated OT's. The speaker cap in the ST120 did work when I overheated it in PA service for 150 people.
Do it yourself means I am in control, not some banker in NYC or Hong Kong. I got probably 2000 hours pleasant low distortion listening out of the modified ST120 before the djoffe feedback bias board failed. What really took it down, I kicked it and pulled some wires out, not internal failure. Maybe next time on the pwgtang circuit I'll get 10000 trouble free hours. The ST70 tube amp sure needs attention every 8000 hours, either caps, output tubes, or rectifier tube. And the vacuum tubes come mostly from Russia, our kind gentle partners in commerce, right? DIY, the parts to update my ST120 come mostly from Mexico, or oriental countries that are not building an island in their neighbor's sea to protect their oil grab.
You bare need to tighten the nuts on your transformer. Elastic stop nuts, maybe, time marches on.
Last edited:
This is the most recent ST120 upgrade thread.
I have a ST120 with a home built djoffe feedback control output transistor bias circuit, This sounds very good with adequate low volume bias current http://www.diyaudio.com/forums/solid-state/156627-dynaco-stereo-120-can-beautiful.html , but has not proved reliable. The match of the two input bias sense transistors can creep off, as I found after a couple of year operation. Due to the fans my ST120 was operating reliably, and sounded great, but one channel had 250 ma of bias current due possibly to the mismatch. (It still passed the double diode meter test) I had at installation blown up a 2n3904 sense transistor with a little extra heat sink grease on the collector connection of one output transistor, so I'm moving on to something else home built. No, I won't be using an LM3886 IC, my unit puts out 80v from PC14, with current regulation set at 6.75 amps, and I like it that way. I also like finned heat sinks, Mr Joffe sells solid 3/8" ones.
So any circuit I install has to cope with 80 v rail, no center tap on the transformer, and a speaker cap. I like the speaker cap, it saved my $125 speakers (Peavey 1210) one time when I overheated the ST120 in PA service and melted the solder on the wire to the output capacitor, causing a shorted OT.
I was installing the pwgtang mod on PC15 which replaces the zener diode, the capacitor coupling of input to drivers, and installs a conventional two diode plus pot bias control circuit. However, a circuit board land between the driver transistors has to be cut. I don't want to do that, the board is butchered enough by twisting around the leads on TO220 transistors to fit the TO5 holes. So I'm going to have to build one PC15 replacement board .
In speaker capacitor circuits with one pair of output transistors (design 2n3055, I've got NTE60, probably white box MJ15003 equivalents), the candidate designs are Armstrong 621 output section, Pwgtang modification of the ST120 PC15, the Apex AX6 "simple retro circuit", The sakis g-amp (for geritol) and the Johm Ellis Basic 50. The Ellis hasn't been discussed since April 2011 (thread 185843-basic 50w amp for beginners) and the discussion then was mostly rambling about geography. The Ellis allegedly has .02% distortion, but I don't think he built one or simulated it. It is designed for 63 v rails. I've got 80. The G-amp has a 63 v LM317HVT regulator and the H package is limited to 0.5 amp passthrough. I don't want to buy a LM317hvK TO3 package from someone that has a $10 freight minimum. ( Newark is $5.50 freight and don't have the TO3 317HVK). I want to listen to records this week, and not order anything elsem I;ve got LM317HVT TO220's, stupid me.
The Basic 50 uses BDxxx transistors, and Mr Ellis talks about how wonderful the BD140 is as a driver with a 100 mhz Ft in a TO126 package. My source of Bd140, newark, stocks fairchild, which has no Ft on their datasheet. Possibly a TIP32b in a TO126 package. I've already got TIP31-2C and TIP41-2C transistors, I won't be ordering anything special. 3 Mhz Ft might hurt HD a lot. I've got MJE15032-33 in stock with 30 mhz Ft at Vceo 250, but at $4.50 each I don't think I will experiment with those.
The Apex retro amp has no bias control pot or control. The Armstrong has a bias pot, is well respected for 1966? but was the starting point for Mr. Ellis's work.
I'm working with self drilled Nema C laminate with wire component to component, to solve the no printer, and no hazardous waste disposal problems ($400 a barrel here) of making circuit boards. I also have had all my Microsoft windows programs destroyed by microsoft by unsolicited updates, so I won't be running any sims or layout program for a commercial PC board maker. I'm using ubuntu linux (until the bash virus gets me, backing up drive this weekend).
The parts count is as follows:
ST120 PC15-Pwgtang -
12 R, 3 e-cap, 2 disc cap, 3 diode, 4 transistor, 1 pot 24 parts
possibly 2 more disc cap to install the TIP mod driver slowers
Armstrong 621
15 R, 4 E-cap, 2 disc cap, 3 diode, 5 transistor, 1 pot, 30 parts.
Apex AX6
19 R, 5 e-cap, 2 disc cap, 2 diode, 4 transistor, 32 parts.
Basic 50
16 R, 4 E-cap, 4 disk cap, 3 diode, 7 transistor 37 parts
All the above leaves out the rail cap, the speaker cap, the zobel, the output network, and the rail fly back diodes. Most are built into the ST120 and the TIP mod I installed put the missing resistor of the zobel on the speaker jacks.
So the pwg tang PC15 wins on parts count and possible lowest wires to solder. It is probably also just as loud as the PC 15 with the djoffe bias control circuit. The St120 was just as loud (gain) as the CS800s with the volume pot hacked up, at 1/3 the power rating. John Ellis spends a lot of text talking about the reduced gain on his input transistor to reduce clipping. The only time I've heard clipping on the ST120 was when I do a power test with a transistor radio input turned up to sound funny- ie clipping all the time. Even in the front yard on 4th of July on the 1812 overture cannon shot, I didn't hear clipping. Nor in the Christmas cantata driven by a keyboard and mono Ampeg mixer, where I melted the solder in the ST120, did I hear any clipping.
So I'm drilling the NEMA 3"x5" laminate card for pwgtang tonight. Any thoughts?
I have a ST120 with a home built djoffe feedback control output transistor bias circuit, This sounds very good with adequate low volume bias current http://www.diyaudio.com/forums/solid-state/156627-dynaco-stereo-120-can-beautiful.html , but has not proved reliable. The match of the two input bias sense transistors can creep off, as I found after a couple of year operation. Due to the fans my ST120 was operating reliably, and sounded great, but one channel had 250 ma of bias current due possibly to the mismatch. (It still passed the double diode meter test) I had at installation blown up a 2n3904 sense transistor with a little extra heat sink grease on the collector connection of one output transistor, so I'm moving on to something else home built. No, I won't be using an LM3886 IC, my unit puts out 80v from PC14, with current regulation set at 6.75 amps, and I like it that way. I also like finned heat sinks, Mr Joffe sells solid 3/8" ones.
So any circuit I install has to cope with 80 v rail, no center tap on the transformer, and a speaker cap. I like the speaker cap, it saved my $125 speakers (Peavey 1210) one time when I overheated the ST120 in PA service and melted the solder on the wire to the output capacitor, causing a shorted OT.
I was installing the pwgtang mod on PC15 which replaces the zener diode, the capacitor coupling of input to drivers, and installs a conventional two diode plus pot bias control circuit. However, a circuit board land between the driver transistors has to be cut. I don't want to do that, the board is butchered enough by twisting around the leads on TO220 transistors to fit the TO5 holes. So I'm going to have to build one PC15 replacement board .
In speaker capacitor circuits with one pair of output transistors (design 2n3055, I've got NTE60, probably white box MJ15003 equivalents), the candidate designs are Armstrong 621 output section, Pwgtang modification of the ST120 PC15, the Apex AX6 "simple retro circuit", The sakis g-amp (for geritol) and the Johm Ellis Basic 50. The Ellis hasn't been discussed since April 2011 (thread 185843-basic 50w amp for beginners) and the discussion then was mostly rambling about geography. The Ellis allegedly has .02% distortion, but I don't think he built one or simulated it. It is designed for 63 v rails. I've got 80. The G-amp has a 63 v LM317HVT regulator and the H package is limited to 0.5 amp passthrough. I don't want to buy a LM317hvK TO3 package from someone that has a $10 freight minimum. ( Newark is $5.50 freight and don't have the TO3 317HVK). I want to listen to records this week, and not order anything elsem I;ve got LM317HVT TO220's, stupid me.
The Basic 50 uses BDxxx transistors, and Mr Ellis talks about how wonderful the BD140 is as a driver with a 100 mhz Ft in a TO126 package. My source of Bd140, newark, stocks fairchild, which has no Ft on their datasheet. Possibly a TIP32b in a TO126 package. I've already got TIP31-2C and TIP41-2C transistors, I won't be ordering anything special. 3 Mhz Ft might hurt HD a lot. I've got MJE15032-33 in stock with 30 mhz Ft at Vceo 250, but at $4.50 each I don't think I will experiment with those.
The Apex retro amp has no bias control pot or control. The Armstrong has a bias pot, is well respected for 1966? but was the starting point for Mr. Ellis's work.
I'm working with self drilled Nema C laminate with wire component to component, to solve the no printer, and no hazardous waste disposal problems ($400 a barrel here) of making circuit boards. I also have had all my Microsoft windows programs destroyed by microsoft by unsolicited updates, so I won't be running any sims or layout program for a commercial PC board maker. I'm using ubuntu linux (until the bash virus gets me, backing up drive this weekend).
The parts count is as follows:
ST120 PC15-Pwgtang -
12 R, 3 e-cap, 2 disc cap, 3 diode, 4 transistor, 1 pot 24 parts
possibly 2 more disc cap to install the TIP mod driver slowers
Armstrong 621
15 R, 4 E-cap, 2 disc cap, 3 diode, 5 transistor, 1 pot, 30 parts.
Apex AX6
19 R, 5 e-cap, 2 disc cap, 2 diode, 4 transistor, 32 parts.
Basic 50
16 R, 4 E-cap, 4 disk cap, 3 diode, 7 transistor 37 parts
All the above leaves out the rail cap, the speaker cap, the zobel, the output network, and the rail fly back diodes. Most are built into the ST120 and the TIP mod I installed put the missing resistor of the zobel on the speaker jacks.
So the pwg tang PC15 wins on parts count and possible lowest wires to solder. It is probably also just as loud as the PC 15 with the djoffe bias control circuit. The St120 was just as loud (gain) as the CS800s with the volume pot hacked up, at 1/3 the power rating. John Ellis spends a lot of text talking about the reduced gain on his input transistor to reduce clipping. The only time I've heard clipping on the ST120 was when I do a power test with a transistor radio input turned up to sound funny- ie clipping all the time. Even in the front yard on 4th of July on the 1812 overture cannon shot, I didn't hear clipping. Nor in the Christmas cantata driven by a keyboard and mono Ampeg mixer, where I melted the solder in the ST120, did I hear any clipping.
So I'm drilling the NEMA 3"x5" laminate card for pwgtang tonight. Any thoughts?
Last edited:
- Status
- Not open for further replies.