Hello SSGA folks, new member here.
My name is Brian and I am a luthier/guitar tech/amp repair tech. I went through a 2 year "electronic music technology" course about 20 years ago here in Red Wing, MN. I also took a guitar repair course and have been a luthier building folk harps for Stoney End Harps for the past 17 years. Recently I've decided to quit the harp business and focus on guitar/amp repair.
My main focus has been on tube amps, but I'd like to get better at repairing solid state devices, as there are not too many folks around my region that do that. Needles to say, I'm a little rusty on my solid state theory, although I have all my notes/text books from school and had good teachers, so I'm hoping to get quickly back in the saddle so to speak. Also, I've always been into tube amps, so I have decent trouble-shooting skills/equipment and know how to solder quite well. I have fixed several pieces of solid state gear over the past several months as well, and have some power amps coming in this week.
Anyway, yesterday I picked up a couple old, not working guitar amps just to mess around with. The one I will ask a question about here is a Heathkit TA-16, and it mainly concerns getting some new parts. So...
1. The power transformer seems to be bad. The 120vac is getting onto the primary, but I only have 3.4vac across the secondary. I know I can't get a direct replacement for this PT...or at least that would not be cost effective at all, so is it possible to find a suitable substitute? How would I go about finding this?
2. As you probably know, these amps came as kits for the customer to complete. This one is put together pretty well, but the bridge rectifier circuit is pretty messy. I think I'd like to replace it with a chip, if possible. What do you think of that idea? Would I be better off just buying the correct diodes and rewiring that way? What chip or diodes would you recommend?
Once I get power on the board, we'll see where I can go from there.
I have accounts with Mouser, MCM and Parts Express, fyi.
I've really enjoyed browsing through the forum and I look forward to being a contributing/learning member.
-BrianS
I got to thinking a little bit and decided to disconnect the primary wires of the PT from the amp and measure that way. Doing this, I get 29vac. The highest DC voltage that the amp uses is 39 volts, so I guess my PT is probably good and I have a shorted component in the power supply?
Also, this amp has circuit breakers instead of fuses, and they seem to be a little suspect. I'm thinking of jumpering across them, one at a time, to see if that helps. Sound too dangerous?
Thanks again!!
QuoteI got to thinking a little bit and decided to disconnect the primary wires of the PT from the amp and measure that way. Doing this, I get 29vac.
What does that mean? :duh
Please draw a diagram of what you are doing.
Quote from: BrianS on June 29, 2014, 01:08:42 PM
I got to thinking a little bit and decided to disconnect the primary wires of the PT from the amp and measure that way. Doing this, I get 29vac.
I think you mean the PT secondary? So when it's connected to the board it reads 3.4VAC, but when disconnected it reads 29VAC?
Is the fuse blowing?
Whoops, typed too fast. I disconnected the SECONDARY wires from the amp!! So, I'm basically just measuring what the PT is doing. 120vac on the primary & 29vac on the secondary.
That makes more sense, right?! Sorry... xP
There are no fuses, just circuit breakers...and they are not popping. I can't be certain that they are functioning correctly however. This amp is in pretty rough shape.
test the diodes in the rectifier and the filter caps.
OK, you must test step by step.
1) You started fine, testing the transformer.
Hope you downloaded the schematics shown in http://www.ssguitar.com/index.php?topic=1411.0
2) now with amp off and unplugged from the wall measure PSU diodes, D102/3/4/5 .
Build a lamp bulb limiter (search SSGuitar) , plug the amp there and turn it on.
I *guess* it will shine quite bright, which means a short :(
Measure voltage across C104 , Q9 and Q10 , in each case CE (collector to emitter) and BE (base to emitter).
Post results.
Remember to include polarity, "+20V" and "-20V" is not "20V".
Two of the diodes in the rectifier were shorted, so I need to rebuild that before I go on.
I do have a bulb limiter and use it religiously. I also have the schematic for the amp in question.
I will post results as soon as I have them...
Way to go. :dbtu:
Try to see if you can "separate" the +38V source (the + end of the big filter cap) from the rest of the circuit.
Don´t know if there´s a wire joining them or you have to cut a track, the idea is checking that you can reach those +38V without load.
Add a 2K2 or 4K7 1W resistor across that cap so it self discharges even without external load.
I suspect you have 1 or 2 dead power transistors, let´s see.
Well, I replaced all the rectifier diodes with a single bridge rectifier "chip"; got rid of the circuit breakers (one was bad) and added a mains fuse and installed a 3 prong power cord. Plugged it into the leak light, which did not light up, and checked my voltages. All were just a few volts off thanks to the leak light.
I also measured for DC voltage across the speaker leads and had around 5 vdc, but that steadily dropped down to less than 3 volts, so I decided to plug a guitar in...and the amp worked! The normal channel works fine and the reverb channel won't work unless you jumper the footswitch jack. I could not get the tremolo to work, and there is no reverb tank, but I did get a dry guitar signal with the jumpered jack. I don't have a footswitch, or a "stereo" plug that the footswitch uses, so I'm totally sure that the tremolo doesn't really work, or if it is just that I don't have the proper plug. More to tinker with I guess.
Once I get the reverb channel working, I think I will add a speaker jack & build a head cabinet for this thing, as the original cab was basically disintegrated when I received it...and I finished the process getting the chassis out :trouble.
I'll post some pics and sound clips if anyone is interested. Thanks for your help!
Oh, I forgot to ask, is there an easy/quick way to determine the correct fuse to use in this amp? Right now I just have a 3A fast blo fuse in there, but I know that is probably too large to do any good.
Is the fuse not on the schematic?
Not in this case, it used circuit breakers.
They must have been on sale, it has a 2.11A one in the secondary which is probably fine, and same 2.11A on the primary which is ridiculous.
It´s an around 25W amp (30W if optimistic) so I guess a slow blow 500mA or 750mA would be fine there.
So T500 or T750 .
Hi
Brian, welcome.
Quote from: BrianSDC voltage across the speaker leads and had around 5 vdc, but that steadily dropped down to less than 3 volts
This is a bit of a worry that needs to be explored. Please
disconnect the speaker until we have this fully explained - it implies that
there is still something wrong with the output stage (perhaps the cause of the rectifier failure?). Output residual DC must be less than half a volt absolute max, normally only around 100mV.
Is it sounding a bit weak and more distorted, not fully clean?
A failed transistor in the output stage may put a high DC voltage and resulting high current on the speaker, cooking its voice coil.
^Thanks. I was pretty sure there shouldn't be much, if any DC on the speaker, but this is not something that I deal with when repairing tube amps!
Yes, the amp distorts fairly quickly...maybe around 5 or less on the volume knob. The distortion is a decent "crunch", though, so I' wasn't sure if this was normal or not.
The amp is fairly noisy, too...I'm sure it needs some new caps and maybe some of the old carbon comp resistors should be replaced.
I will test the output transistors today...
Both output transistors test bad using the ohm setting on my DMM. One seems to test "worse" than the other, but that's kind of speculation on my part. The PNP transistor fails all tests: Emitter to base, base to collector, Emitter to collector, reversing the leads...I get conduction with all these tests. The NPN transistor passed the emitter to base and base to collector test, but not the emitter/collector test.
Guess I need to hunt down some new transistors...
Protip: remove and test both the driver transistors for the output pair. With s.s. power amps whenever you find a dead semicon, particularly the very common dead output one(s), you really need to look at its driver as well. If either of these are dead go for a snoop back to the Voltage Amplifier Stage/bias network, protection transistors, and so on.
Normally what happens when an output transistor fails is that the amp suddenly starts producing a very loud mains hum, it is generally switched off in panic within a few seconds, and this generally saves the rectifier (and power transformer), but not the reccy in this case, so it may have been flogged or left on for some time post-failure and we have to be more alert for other collateral damage, e.g. a driver transistor or two, the occasional roasted resistor, &c.
You must use the limiting lamp until you are quite certain that the output stage is again re-balancing the output half-rail to zero (or very near zero) volts across the speaker connection. Together with only a dim lamp this balanced condition is a pretty fair indicator that the power amp section is working as it should.
Being DC and direct coupled s.s output stages look more like a hairy-legged op-amp than a valve output stage, and they are less forgiving of mistakes, but the use of a limiting lamp is the main difference in servicing. Unlike a valve amp the output of a s.s. under service should be left open until you measure less than half a volt on the output. There may be some other tips but really it's just more electronics (and you already know how to deal with that around valves, right?)
There is some physical stuff about insulating washers when mounting semicons on heatsinks (and sometimes live heatsinks, e.g. a fat Ashdown I worked on).
A while back I knocked this page up to help people deal with s.s. amps (particularly those who may already have some valve experience);
http://www.ozvalveamps.org/repairs/solidstateamprepair.htm (http://www.ozvalveamps.org/repairs/solidstateamprepair.htm)
HTH
Pity this thread was called new member and a question which means absolutely nothing.
I don't even remember what amp are we talking about and it definitely will not show in a future search for somebody looking to repair a similar one.
It would have been much better if it were labelled : New member and a problem with xxxyyy amp or whatever.
In fact you can skip the new member bit ... we *all* were one once.
Just sayin'
Roly, Please look at the schematic. This amp is a single rail design with an output capacitor. The 3V on the output is just leakage from the output cap.
Brian, check the voltages around those output transistors and compare to the schematic.
Quote from: Loudthud on July 03, 2014, 11:50:03 AM
Roly, Please look at the schematic.
Circuit? {rummagerummage} Circuit??? {rummagerummagerummage}
Ahh ...
there it is ...
http://ctgelectronics.weebly.com/uploads/3/1/6/6/3166248/ta-16_schematic.pdf (http://ctgelectronics.weebly.com/uploads/3/1/6/6/3166248/ta-16_schematic.pdf)
:tu:
Oh, it's only a baby; yeah a new OP pair should fix that (but I wonder what the basic cause was?).
With that level of leakage the OP coupling cap might be worth a close look as well.
I'm worried about those output transistors. My data says the 2N2148 is Germanium rated at 12.5W. It won't look good on most DVM diode test functions. I have no idea on the TA2577. There isn't much drive current available to Q10 so it needs high beta.
Quote from: LoudthudGermanium
Oh oh. Well spotted. :dbtu:
{We really must get that repair form thingy going so (I'm) on the same page. ::) }
QuoteI'm worried about those output transistors. My data says the 2N2148 is Germanium rated at 12.5W. It won't look good on most DVM diode test functions. I have no idea on the TA2577. There isn't much drive current available to Q10 so it needs high beta.
I tested both transistors using a basic "transistor test" using the ohms function on my DMM...maybe there is a better way to test these. Also, I seem to remember learning something about "leaky" transistors, meaning that they sort of work, but "leak" current where they're not supposed to? Could these transistors be "leaky", but still function somewhat? I mean, I know they function, as I played my guitar through the amp and actually got some decent tones out of it...
QuotePity this thread was called new member and a question which means absolutely nothing.
Sorry about the title faux pas. Some forums expect a formal introduction from new members before questions get answered. I didn't see any kind of rule or introduction page, so I just went with what I thought would be the most effective. I also searched this forum before I joined, and found several references to the TA-16 in a number of different threads, so I assume that this thread will pop up if someone searches the right terms.
If there is a moderator that could edit the title for better search-ability, that would be fine with me.
Thanks to all that are participating in this thread.
I'm off to measure voltages...
DC voltages on the output transistors:
Q9: Base: 19.15
Emitter: 19.3
Collector: 0 (connected to ground)
Q10: Base: 19.83
Emitter: 19.3
Collector: 33.8 (This is my "B+" voltage (rail voltage?))
So, my main "B+" voltage is about 4 volts less than the schematic, but my base and emitter voltages are about 1 volt higher than what is listed on the schematic. However, the voltage drops across the base/emitter junctions are just about spot on with the schematic.
If this was a tube amp, I'd say these parts are functioning as they should!
Oh, the amp is also still plugged into the limiting bulb, so I suppose that is affecting the B+ voltage. Is the correct term "rail voltage" here? There is no negative rail, so...
When testing germanium devices (which we hardly ever see these days) there are two significant differences to the common silicon devices we normally deal with.
The forward cut-in voltage of a silicon junction is around 600mV +/-100mV, depending. The same voltage for a germanium junction is about 100mV. If we test a germanium device thinking it's a silicon device we would then say "ooh, shorted junction" but it isn't - it's normal for germanium.
The second trap is leakage. A silicon junction in reverse, or transistor between Collector and Emitter with no Base injection, is as near an open circuit as makes no never mind. A germanium junction or transistor in contrast normally has some leakage when its supposed to be off; that as healthy as they get.
This makes testing germanium devices a bit harder, but in your case if these devices are giving reasonable transistor action in circuit, that is they seem to be amplifying the current between the input and output, then that is a good enough test.
{yeah, "rail" is a bit of a buzzword, but it has been around since at least the 1920's when perhaps power was distributed around equipment by large exposed metal bars mounted on elegant insulators.}
Quote from: BrianS on July 03, 2014, 04:26:34 PM
DC voltages on the output transistors:
Q9: Base: 19.15
Emitter: 19.3
Collector: 0 (connected to ground)
Q10: Base: 19.83
Emitter: 19.3
Collector: 33.8 (This is my "B+" voltage (rail voltage?))
So, my main "B+" voltage is about 4 volts less than the schematic, but my base and emitter voltages are about 1 volt higher than what is listed on the schematic. However, the voltage drops across the base/emitter junctions are just about spot on with the schematic.
If this was a tube amp, I'd say these parts are functioning as they should!
Oh, the amp is also still plugged into the limiting bulb, so I suppose that is affecting the B+ voltage. Is the correct term "rail voltage" here? There is no negative rail, so...
Looks good :tu:
If the top transistor were shorted or horribly leaky, the mid rail would be practically +B , +38V or whatever.
If the bottom one were so, you would have close to 0V
But you have roughly half +B there, so they are proving themselves fine without further testing.
As of voltages on schematics, usually 10% away is fine and 20% away acceptable, so no problem there beither.
Now what was the remaining problem?
Thanks for all the advice thus far, folks! Here's a couple recordings of the amp. The first one is the amp with no production effects. I do turn on an OCD to give the amp more treble, as it is a little "dull" on its own. The recording is rather long, so skip around to hear various tones. I get to the overdriven tones near the end:
https://soundcloud.com/brianstewart-1/heathkit-ta-16-dry (https://soundcloud.com/brianstewart-1/heathkit-ta-16-dry)
Here's a recording with some reverb/compression/somethingelse that I put on with my DAW, just for fun:
https://soundcloud.com/brianstewart-1/heathkit-ta-16-with-effects (https://soundcloud.com/brianstewart-1/heathkit-ta-16-with-effects)
(https://lh6.googleusercontent.com/-3kJ5aP8oupk/U7cG5zePZ1I/AAAAAAAABhU/KPeYzVOxBRk/s800/008.JPG)
QuoteNow what was the remaining problem?
Well, there was dc on the output, so replaced that coupling cap, which solved that problem:
(https://lh4.googleusercontent.com/-h8ZZ4uhxYRE/U7cG5_BNu0I/AAAAAAAABhg/j7xrL3PIDUs/s512/004.JPG)
The main filter cap had a bulge in the end, so I replaced that as well. I used what Radio Shack had on hand, which was (2) 2200ufd, 50v caps that I wired in parallel. The original was 4000ufd, 50v:
(https://lh5.googleusercontent.com/-rwhIROMX_7M/U7cG52-m1DI/AAAAAAAABhc/FOyhg6x2HzM/s640/005.JPG)
That pic also shows the new bridge recto and the new 3 prong cord.
Now, the only things left are the tremolo and reverb. I measured the voltages on all the remaining transistors, and they are good, so my guess is that the opto-coupler is bad in the trem. circuit. Internet sleuthing revealed a good replacement, which I will order soon.
There is no reverb tank, and I really don't want to spend the money on one yet, so we'll see what comes of that. I also need to figure out the correct tank to get. I have to match the input and output impedance of the circuit, correct? Need advice here.
I also will be building a head cabinet for this at some point, as I have nothing else to put it in.
Bridge rectifiers like what you installed are typically only rated at 1 Amp. That's a little too low. See if you can find one rated at 4 Amps or more. For voltage rating, anything over 50V is fine.
I had a heatsink with a 2N3055 and MJ2955 mounted so I just sky-wired the rest of the output stage, Q8, Q9, Q10. The beta of the 2N3055 was too low, the amp only made about 20W at 4 Ohms and about 18W into 8 Ohms. Your output stage seems to be working so finding replacement transistors won't be a problem.
Quote from: BrianSWell, there was dc on the output, so replaced that coupling cap, which solved that problem:
Ah! :dbtu:
Quote from: BrianSI also need to figure out the correct tank to get. I have to match the input and output impedance of the circuit, correct? Need advice here.
Some things are easier than others. The circuit shows the reverb unit as "4FB3A1B". I'l make a fairly large bet that this is an Accutronics/Belton part number and can also be decoded to find the tank specs. <http://www.accutronicsreverb.com/ (http://www.accutronicsreverb.com/)>
4 = 04 series
F = input 1475 ohms
B = output 2250 ohms
3 = long decay
A = input and output grounded
1 = no transport lock
B = mounting horizontal open side down
Just search "4FB3A1B" for offers, ~$20 +shipping.
HTH
And yeah, 1 amp diodes are a bit light-on, 4 or 5 amp would be safer (and don't worry about excess voltage rating, that's fine).
Quote from: LoudthudThe beta of the 2N3055 was too low
This is their classic problem. For all their hairy-legged ratings, e.g. "Ic
max = 15 amps", the
Hfe drops to about 20 at only Ic = 5 amps. Practically this means that you really can only get about 50 watts into 8 ohms out of a pair (on +/-36V), and that a pair alone won't be happy on 4 ohms loads. There are far better transistors available these days, but they don't often turn up in the "throwout" bin by the door for 20c each. :cheesy: The homebrew Twin-50 I use as my stage synth amp (PA, foldback, etc) uses a pair in each channel and they have served well.
Well, 3055/2955 used *without* drivers in that simple circuit will definitely lack Hfe.
Add a couple Tip 31/32 which multiply it by at least 40 and now you have an easy to drive compound pair, still 15A rated and very usable at 10A.
Roughly same thing comes neatly packaged as TIP142/147 , with guaranteed Hfe of 1000 at 10A :o
That's why it's practically the new standard industry workhorse, used by none less than Fender, Marshall (in an European incarnation), Laney, Crate, H&K and tons others (including yours truly).
EDIT: got real curious about what speakers were used there, got a couple pictures?
Would love to see what frames, magnets, cones they used.
Also the frames might have some small rubber stamped numeric codes , such as 2207614P12Q which although looking very mysterious actually means easy to decode Jensen week 14 of 1976 model P12Q .
Yours might show a similar one.
Sorry to nitpick, but standard "safe" wiring of fuseholder is for incoming power to go to back end of fuseholder first, then out the side. This is so as soon as the fusecap is disconnected the hot can no longer touch anything. With the hot on the side terminal like you have it, as the fuse is being pulled out it can contact the "hot" side terminal.
Of course we assume the unit will be unplugged before the fuse is removed, but mistakes happen.
QuoteSorry to nitpick...
No problem g1, thanks for pointing that out! I will change it when I redo the recto and replace the optocoupler.
Quotegot real curious about what speakers were used there, got a couple pictures
If you are talking about my recordings, the speaker is an 8ohm Eminence Legend that I bought used. Model designation is AVM-128.
I purchased the TA-16 sans speakers.
Quote from: J M Faheyvery usable at 10A.
That depends on what you mean by "usable".
QuoteVce(sat)min = 3V @ Ic = 10A and Ib = 3.3A
Over
three amps of Base current - for only 10 amps Collector current? They don't give any figures for the maximum current rating so we can only guess how bad they are. Need another '3055 as the driver.
At Ic = 15A Hfe is around 10. (all the
ON Semi datasheet plots actually end at 10 amps, well below Ic
max which I think contains a message)
I love the '3055 like a brother, but ya gotta be realistic - these days there are far better devices at reasonable cost.
QuoteQuote from: J M Fahey
very usable at 10A.
That depends on what you mean by "usable".
I can get 10 A peak into a load with only 20 mA drive ... and that´s the "min" Factory guaranteed spec, I design with 10mA drive and it´s always enough.
Practical use: a +/-40V rails power amp with TIP141/142 outputs can put full rail voltage (minus transistor drop, some 4V and what you lose in the ballast resistors) into a 4 ohm load.
That´s why TIP142/147 are *so* popular in major brands.
http://www.onsemi.com/pub/Collateral/TIP140-D.PDF
Sorry JM, I think we've got crossed wires. I'm saying that the 2N3055 isn't worth a crumpet above about Ic = 5 amps.
Yes, if I was looking at more than 50W or 4 ohm loads I'd be considering something else, such as the TIP 142.