Hartke HA5500 Bass Amp

[CJ]

now what? another amp thread?

well, this one sounds good, so WTF, over?

Billy Sheehan and Will Lee like it, supposedly,

or they could just be sayin that to get free amps, 

supposed to be 500 watts,

notice the busted pot shaft which will require major surgery to fix,

 

[CJ]

back panel layed out nicely,

you can pull that little hook and find fuses along with a 120/220 jumper thingy,

 

[CJ]

forced air cooling thru a square box, and another fan on the transformer,

kind of loud, two fans will create beat frequencies which double the sound,

de-populated board means a stereo version might be an option,

there is a tube under that metal thingy, so you have two preamp options, tube or opamp, a compressor, master vol, graphic,

 

[CJ]

12AX7a, Analog Devices, JRC...

 

[CJ]

parts is parts,

 

[CJ]

how many cable connectors do we get to pull in order to change that 100K linear pot?

wish it was a Fender, would be done by now,

they should make the engineers work on their own stuff,  ;D

 

[CJ]

power board, top FETs are easy but there is another row burried next to the chassis bottom plate,

you can but this whole board for about $129,

 

[CJ]

schemo>

http://www.eserviceinfo.com/download.php?fileid=47119

 

[PRR]

> schemo

Value of R315 is surely wrong, 100X low. 68K??

Ah, someone put R315's value at R316.

 

[CJ]

sometime they put in the wrong value to trip up the bootleggers,

getting the 100K linear pot out of this thing, first, panel #1 comes off,

if this was a Fender i would be done by now,

 

[CJ]

took off the rest of the hardware, no room to slide the PC board back so the pot shafts clear panel #2, 

so off comes panel#2 

if this as a blackface Fender i would be done by now,

 

[CJ]

back side of board, pots have a different look to the solder, means they were attached after the board came out of the wave tank,

there are about 3,641 linear pots listed in the DigiKey pages,

got to get the right size, right resistance, right shaft dia and length, got to get the split shaft with splines, side adjust, top adjust, rear adjust, detent at 12 o'clock, time consuming,

https://www.bourns.com/data/global/pdfs/PDB18.pdf

so we order the PDB181-K220K104B from DigiKey for $1.26 ea, $2.86 shipping,

then you have to find a knob because the old one is still attached to the broken pot shaft, probably sitting in an ally behind Studio 54 somewhere or maybe CBGB's,

so you scour the net for another marathon session, sorting thru 10,000 control knobs,

close, but no match, BassTalk forum says go to radio shack,  :'(

so you call all the Hartke authorized service centers and they tell you to call Hartke,  :'(

so you call Hartke Service in NY and the guy says he will send one out for free, 

1-631-784-220

you know something.....



















if this was a pre CBS Fender,  i would be done by now,  ;D

 

[ricothetroll]

Hi,

I happen to have that same exact amp on my bench. It's here because the fuse blowed a few seconds after the amp is switched on. I have disconnected the supply of the amp board to that made me conclude the issue is on that board (and not on the PSU). Visually, nothing looked burnt or damaged so far.

I the disconneted all 8 power transistors to check for continuity or shorts, they all looked fine.

Then I had this brilliant idea : why not power the amp board with the power transistors out to see if the fuse still blows ? By doing that I also thought it wasn't necessary to connect the signal connector, not seing that it provides a different ground to the board (both grounds are actually linked through a 10R resistor). Power on... BAM ! Q310, the lower driver transistor, also connected to the heatsink, explodes ! I imediately switched the power OFF, and notice that R304 also burnt.

Looks like it's time to recognize that this one is (at least) a little bit too complicated for my own skills and it's time to get some help to understand the details of that schematic, both for my own education and to avoid doing other mistakes....

What I can say so far :

- Q301A/B, along with the current source/mirror Q302/303, form a differential amp. It takes the input signal and applies some NFB through R318 from the power output

- Q306/307 forms another current mirror/source, I don't get what it's used for

- Q308 and the surrounding components form the push/pull biasing network, seems temperature compensated by TH302

- IC301 protects the speaker and controls the output relay

- Q311 and Q312 look like they stabilise the biasing, though I don't understand how very well. Q311 can only draw DC through Q305, and Q312 through Q307. Their interaction is mysterious though

- The part I understand the least is Q304/Q305 : the arrangement reminds me of a diff amp but I don't get what are the signals involved : why is the collector of Q305 connected to the driver's base (Q309) ? What's the role of this arrangement ?

Thanx in advance !

Best regards.

Eric

PS : schematic here https://drive.google.com/open?id=0B6K-0QVHN7zUbm9DXzAzdWhXWGM

 

[ricothetroll]

Hi again,

Looking at that schematic again and again I'm starting to have a clearer view on what's happening in there.

So :

- Q301-A/B is, biased with current source Q302/303, the input diffenrential amp, which substracts the feedback from the global output to the input signal, thus forming a NFB circuit.

- Q304/Q305 form another differential amp, biased by R304 and loaded by current mirror Q305/Q306. This provides a huge amount of gain, that's in the end controlled by the NFB. The power amp as a whole then could be seen as a BIG opamp.

The rest is pretty classic, with Q308 and surrounding components an adjustable/temp compensated push-pull stage bias circuit, Q309/Q310 a pre power transistor driver and Q313>320 the output transistors. The rest is protection and Q311/Q312 form some kind of output limiter.

That being said, the amp is still not working :

- I have tested the power transistors, all good.
- The drivers blew when I powered the amp with the power transistors removed
- R304, the second differential amp's bias resistor (68R), blew also.

So, my interrogations are :

- How come did R304 and the driver blew whan I tested the amp without the power transistors ?

- How can I test the circuit before the drivers without the output transistors (and thus without NFB) ? The huge gain provided by the current mirror loaded second diff amp will lead me to over-saturated results for sure. Of course, I'd have replaced the burnt R304.

- I then need to replace the drivers. As KTC4370A / KTA1659A doesn't seem to be available anymore, I need an equivalent. Performing a parametric search on Farnell, I could find some candidates that respect the max specs, but those are always much more powerful transistors, like 150W instead of 15W. What's the pitfall with using such a replacement ? What are the most important parameters in that situation, besides the max specs ?

Thanx in advance for your help.

Best regards.

Eric

 

[CJ]

check the AC voltage coming from the toroid pwr transformer, sometimes they wind them with too may turns and thus you get too high a voltage feeding the output circuit,

 

[PRR]

Assume negative rail is -55V.

This splits over R315 Q303 R312 to put about 50V across R312. 50V/10K is 5mA.

Q302 R311 also pass 5mA.

Diff-pair splits this, 2.5mA each side.

R302 R303 1K with 2.5mA will drop 2.5V.

Q304 Q305 B-E drops 0.6v, so 1.9V across R304.

R304 should dissipate 0.053 Watts. It should not smoke.

Dumb-guess is that Q302 is passing WAY too much current.

Hard-knocks guess is that something ELSE has been melted into a dead-short.

First, use ohmmeter and look for a dead-short on any transistor. Depending on your meter current, you may get few-ohm or high-ohm one way, infinity the other way, but never sub-OHM both ways on two pins.

I would toss Q309 Q310 without even looking. When an output transistor fails, the driver transistor "tries" to pull the load. This never ends well.

Cordell's book will teach you all you need to know about design details like Q306 Q307. It is heavy meat, I been chewing for months. He does not go into how to FIX amps, which is rather a special knack. (And a costly hobby.)

 

[ricothetroll]

Hi,

Thanx a lot for your answers ! Those has been really helpful for both understanding and (hopefully) troubleshooting the issue.

I did find a dead short in Q305 ! My interpretation is it sent Vcc (around 100V IIRC) directly through R304, that burnt it at least partially. It also biased the drivers base nearly 100V, and the base of the power transistors. The top ones, with 100V at the base and near 0V at the emitter turned fully ON and thus blew the fuse. When I removed them, the fuse didn't blow anymore but there was about 100V at Q310 base and -100V at its emitter, that's why the poor bastard exploded.

So, I have at least another KTA1659A to replace...

Any thought about a replacement ?

God I love this place.

Best regards.

Eric

 

[walter]

Assume negative rail is -55V.

This splits over R315 Q303 R312 to put about 50V across R312. 50V/10K is 5mA.

If so C305 may be under rated at 50v...

 

[ricothetroll]

Hi,

Well, some progresses have been done but I'm still not there yet...

I've changed Q304, Q305 (voltage gain, second diff amp), Q309 and Q310 (drivers) :

- first with some MJE15032G/MJE15033G because they could stand the voltages / current necessary for the circuit. Yet the output was fine without a load but with a load I had some very early clipping and crossover distortion. I thought this was because of the lower hfe compared to the original KTA1659A/KTC4370A : drivers needing too much current on their base and overloading the previous diffa amp.

-  I was then confident in ordering the "real" replacement needed. I finally got them this afternoon, threw them in the circuit and... the same ! The output is still perfect without any load, but with 16R connected to the output, I have gross distortion (gets worse with 8R/  4R).

A few remarks :
- I've only tested with a resistive dummy load, not with a speaker yet, but I don't think this would change anything.

- That circuit is pretty difficult to debug because it' s like a big opamp : huge amount of open loop gain with negative feedback from the output. Makes signal tracing difficult because anywhere the distortion takes place, the whole circuit has distorted signal in it...

What could cause such a sensibility to load ? My interpretation worked well on the paper, but it seems that something else is going on....

Thanx in advance for any advice !

Best regards.

Eric

 

[CJ]

i have had problems when using resistor loads on transistor amps, do not know why, try a speaker box,

also, hooking a scope up can cause problems, are you using an iso xfmr for the scope/

try reversing the ground and hot lead,