Kustom amp question.

Started by digi2t, October 21, 2013, 01:55:51 PM

Previous topic - Next topic

R.G.

Quote from: PRR on October 22, 2013, 02:36:18 AM
4,500uFd 50V? Bah, that's not big anymore. 'Bout a 1-inch can. 3 bux.
Typically *about* --
Height 1.378" (35.00mm)
Dia 0.866" (22.00mm)
Second that. I have found a number of surplus caps that would be an upgrade for the amp. I have a batch of 18,000uF/63V snap-ins that work nicely in amps around this power. They're about 1.375" diameter by 2" tall. Cost about $1.50.

The biggest problem is mounting them. Cap clamps are a bit silly for them. I hacked small rectangle of PCB stock, hand drew and etched the pattern, and used the PCB stock on standoffs to mount a pair.

Be sure to use snap-ins. They have the ESR and ripple current specs you'll need. The smaller wire-lead stuff is iffy because they may overheat from the ripple current.

Paul sez:
QuoteThis power-amp lacks any coil or R-C at the speaker terminals. "Most" basic transistor amps "require" these. I wonder how wild the supersonic response is?

Also note that the "protection" transistors only sense current, not voltage (or SOA). Further: the averaging network means that when Q10 fails open, Q11 is allowed *double* the total current (four times its nominal protection-current) and ain't going to last another loud note.
Yep, I'd add a Zobel and a damped inductor in series. And tinker a bit with the  protection. If you want, I'll hack on the schemo a bit.

And there are other things. Mark sez:
QuoteThe ones in the power supply will pose a challenge for replacing, largely because any modern replacement will be about 1/5 the size and not fit in the harness provided.  That will certainly make it a lighter amp, but you'll still need to figure out how to secure them.  I'm sure you'll work up something clever.
And he's right.  I recommend an idiot PCB and snap ins. And the harness is another issue. The harness shown will give you excess hum. I just noticed that there is only one wire going into the joined middle of the power supply caps. There should be two; one from the rectifiers and one from the circuit. The joined ground should be a metal strap with the in and out wires at the geometric middle. This minimizes the ground shifts from rectifier current pulses.

R.G.

In response to the questions in the forum - PCB Layout for Musical Effects is available from The Book Patch. Search "PCB Layout" and it ought to appear.

R.G.

And the power amp schemo is drawn to be as confusing as possible.   :icon_biggrin:
R.G.

In response to the questions in the forum - PCB Layout for Musical Effects is available from The Book Patch. Search "PCB Layout" and it ought to appear.

teemuk

QuoteSo, it would appear to have a kind of treble boost/cut function, but not standard.

It's a standard mid-range control circuit derived from James/Baxandall -style circuit.

Dialed to "5" such circuit has a flat response. On other ends of the dial there's either 20 dB  boost or cut centered around 620 Hz.

QuoteThe ganged 10k section would appear to adjust attenuation, conceivably to compensate for whatever the tonal adjustment does to signal level.

The attenuator coupled to aforementioned circuit has following effect:
- Dialled to "max. volume": bass and treble boosted, mid-range attenuated
- Dialled to center of the dial: flat response, no attenuation
- Dialled to "min. volume": bass and treble attenuated, mid-range boosted


QuoteIt's as if there is some kind of interactive conflict between the preamps, and the voicing circuit.

There's not. You are simply hearing the "voicing" circuit being what it is: a tone control combined to an attenuator.

QuoteIf the treble controls are set too high, a high pitch whistle sets in.

This is oscillation. It's not supposed to take place.

digi2t

#23
QuoteYep, I'd add a Zobel and a damped inductor in series. And tinker a bit with the  protection. If you want, I'll hack on the schemo a bit.

Yes please, if you would sir. Like I said, this amp, despite the aforementioned technical shortcomings, has potential. It also provides a bit of a platform for me to work with, and gain a bit of amp experience, which is almost nil.

Would the MJ15003's, with the Toshiba 2SA1837 / 2SC4793 be OK in this amp? I really can't find any data on the original RCA transistors in this unit. These were recommended by you R.G., a while back, and they worked smashingly well in my Ampeg. To my untrained eye, looking at voltages, wattage, and whatnot, it seems to me like they would do. I figure if I'm going to recap, I might as well beef up the power end as well, as I did the Ampeg. Small price to pay for many more years of service.

Any opinions on the 4558 opamps? Leave them, change them? Let's hear it. This amp has struck a heart nerve with me, so I'm going to upgrade whatever may improve it's performance.

Teemuk; thanks for chirping in here. The way you described the voicing circuit operation is spot on. That's the way I hear it, you accurately wrote what I heard (if that makes any sense  :icon_mrgreen:).

You're correct again on the oscillation. It ranges from just below dog whistle frequency, to up around something that drives my dog insane. :icon_lol: I realize it's oscillation, but I'm using the term "whistle" as a layman term.

EDIT: Many of the polarized caps in here are not cans, but rather, they look like black jellybeans. I'm assuming that they are tantalums, some are up to 100uF/20V. Change them, or leave them?

  • SUPPORTER
Dead End FX
http://www.deadendfx.com/

Asian Icemen rise again...
http://www.soundclick.com/bands/default.cfm?bandID=903467

"My ears don't distinguish good from great.  It's a blessing, really." EBK

UKToecutter

I'm not aware of Tantalum caps degrading in the same way as aluminium electrolytic caps but I've seen plenty go bang!!
That said, I'm sure they would degrade in some way with age.

Must go and read up on that
ShumannPLL BOM
Reserve Boards

Mark Hammer

Change the electros, and maybe the input transistors (if they're hissy***), and take it from there.  Those black licorice things are essentially sealed, so I don't see them getting dry or whatever with age.  But let need determine that choice, rather than mere received wisdom.

***The Heathkit TA-16 I picked up a few years ago was quite hissy.  I read somewhere that the 2N3391s it used declined with age (though I'll be damned if I understand why), so I swapped them out for 2N5089s and the hiss went away like magic.  Granted, the TA-16 was about a decade or so older than the Kustom.

R.G.

Quote from: Mark Hammer on October 22, 2013, 04:12:41 PM
I read somewhere that the 2N3391s it used declined with age (though I'll be damned if I understand why), so I swapped them out for 2N5089s and the hiss went away like magic. 
Base-emitter breakdowns. Reverse-break the base-emitter of a bipolar *even once* and it gets hissier. Many circuits are designed with capacitances inside them so that at power up or down, the base emitter may be reverse biased for a few milliseconds - or microseconds. Each power cycle in a circuit like this makes the transistors noisier.

There are some defensive moves, like putting a reverse-polarized diode across the base-emitter to clamp the reverse voltage, but these also add capacitance and leakage - which may not be a problem. It's like the old "he'll beat you when he's sober - so you don't have to worry."
:icon_lol:

The elegant course is to design the circuit so it doesn't reverse break the junctions. This may not be possible with input transistors where the input can be anything between a short circuit and line voltage, and also kilovolt static sparks.
R.G.

In response to the questions in the forum - PCB Layout for Musical Effects is available from The Book Patch. Search "PCB Layout" and it ought to appear.

PRR

> two big power caps. Giants to me

Yup, that's what 4,500u 50V used to look like.

We called them "Computer Grade" back in the day, but the quality was NOT that good. I'd take any Japan-cap over those any day. Japan has moved out of the cap-racket; China-caps *from reputable vendor* seem to have picked up all the good habits of the Japanese.

Power filter caps didn't get much bigger than that. Too many peas in one pod spoil the soup, or something.

The smaller ones today ARE better. The old ones were a lot of excess aluminum and paper and parasitics.

Consider "Snap-in" caps. Get some sturdy board; Hoffman Amps has too-good stuff by the inch. One edge, drill the 4 holes for the big-cap screws. (May as well straighten them out now.) Trace the "V" on the bottom of the board. Now plan 4 holes to mount the two snap-caps hanging down, with snap-legs up through the board. Run three pieces of #16 wire across the holes and crimp the snap-legs over, solder. (3 pieces so you can catch a + and a - under one wire to be your common ground.) You could mount new screws and leave the harness intact, or snip the ring-terminals and just solder harness to cap pins. Since there's a lot of harness (and Kustom used wire-ties), another couple holes in the board and some new wire-ties to take the shake off the harness before it rips off the cap legs.

R.G. has a similar plan if you like etchant.

> worthwhile to replace the 4558's

If you want a new amp, just go buy one.

This is VERY 1976. Not a bad year (what little I remember). Some people remember it fondly (what little they remember). Would you rebuild a 1952 Rolls-Royce with a Honda engine? Re-write Jesus wandering the wilderness with a GPS? At least try it the way it was built.

I also would not tamper the power-amp a BIT. R.G. would. What-ever.
  • SUPPORTER

digi2t

QuoteConsider "Snap-in" caps. Get some sturdy board; Hoffman Amps has too-good stuff by the inch. One edge, drill the 4 holes for the big-cap screws. (May as well straighten them out now.) Trace the "V" on the bottom of the board. Now plan 4 holes to mount the two snap-caps hanging down, with snap-legs up through the board. Run three pieces of #16 wire across the holes and crimp the snap-legs over, solder. (3 pieces so you can catch a + and a - under one wire to be your common ground.) You could mount new screws and leave the harness intact, or snip the ring-terminals and just solder harness to cap pins. Since there's a lot of harness (and Kustom used wire-ties), another couple holes in the board and some new wire-ties to take the shake off the harness before it rips off the cap legs.

I've done some research over night, and I'm cool with the snap in caps. Not cool with the "hanging package" idea though, if that's what you're suggesting (that's what I'm visualizing, in any case). I would prefer to keep the same configuration with the new caps, replacing the clips with proper size ones, of course. As for the screw terminals, they really don't appeal to me, and would prefer the clips, and go with solder. I was thinking of using a piece of single sided copper clad, cut in a strip and drilled, to act as a buss bar between the + and -. I would drill holes in the exact center to solder the wires between the two poles. This should meet R.G. recommendation concerning rectifier current pulses, I think.

QuoteIf you want a new amp, just go buy one.

This is VERY 1976. Not a bad year (what little I remember). Some people remember it fondly (what little they remember). Would you rebuild a 1952 Rolls-Royce with a Honda engine? Re-write Jesus wandering the wilderness with a GPS? At least try it the way it was built.

Yup, 76 was a great year. We had the Olympic Games here, and we spent 3 wonderful weeks in Florida during America's bicentennial. Great year.

It's not that it doesn't sound great with the 4558's, it's more me being inquisitive than anything else. I ended up swapping op amps in my Ampeg, and I found that it made an improvement (to my ears, where it counts). You are reasonable though, I should wait to see how it sounds after the caps are replaced, and evaluate it thoroughly before I noodle with it too much.

QuoteI also would not tamper the power-amp a BIT. R.G. would. What-ever.

Far be it for me to melee with giants, so I'm going to steer clear of this one.  :icon_rolleyes: I simply wanted to know if there was any way to beef up the power section a bit. I just don't want to live through that dreadful moment again, when my Ampeg was rocking out one minute, and then.... nothing.

If a Zobel network would help, why not? Looked it up on the net, and checking the schematic, I saw that my Ampeg is equipped with one.
It's also for this reason that I asked if MJ15003's, with the Toshiba 2SA1837 / 2SC4793 drivers be suitable in here. They work so well in my Ampeg.

Not trying to put a V8 in a Honda, just beefier shocks on a Jeep. :icon_mrgreen:

  • SUPPORTER
Dead End FX
http://www.deadendfx.com/

Asian Icemen rise again...
http://www.soundclick.com/bands/default.cfm?bandID=903467

"My ears don't distinguish good from great.  It's a blessing, really." EBK

teemuk

#29
QuoteI simply wanted to know if there was any way to beef up the power section a bit.

Aside from mechanic design issues like improved ventilation and heatsinking there's plenty one can do to "beef up":
- Beefier drivers and output transistors. These have to be implemented with care since the circuit itself is designed with specific devices in mind and other ones may cause issues (e.g. oscillation). If neccessary, modifications to circuit must be made to compensate using different devices. No, just because it worked in a specific Ampeg doesn't mean it will automatically work with this amp or others. It may do so, but it may not as well. Research.
- Improved bias circuit with better thermal tracking. The current one is probably biased quite cold to be on the safer side. Today the series diode design is considered antique and it basically works flawlessly only in integrated circuits.
- Zobels are a decent addition. They work above audio range so they are effectively transparent tone wise.
- Catch diodes from power amp output to power supply rails
- Antiparallel pair of diodes connected between bases of the LTP transistors
- Antiparallel pair of diodes connected in parallel with C5. If C5 is not non-polar then replace it with non-polar cap of the same value.
- Baker clamping the voltage amp stage
- Diode limiter to input, threshold configured about 2x higher than power amp's input sensitivity
- Addition of more effective short circuit protections (to complement the VI limiter): Sustained high current draw at output either
  a) mutes power amp input
  b) disconnects speaker load
- delayed mute on startup, undelayed mute on shutoff
- DC voltage protection for speaker load


How worthwhile these beefing options are (especially considering that the designer already put decent effort in making a solid, reliable design and that the amp as a proof of it has lived 37 years so far) is up to the guy who will build them, but these additions do "beef up" the design, unquestionably.

Mark Hammer

If a solid state amp were not desirable or useful for the cleans it produced, I personnally would not bother with trying op-amps beyond the usual 4558.  However, where nice clean highs ARE attainable (and the speakers can support it), trying some bi-FETs or other "better quality" op-amps is not a completely foolish idea.

digi2t

Quote from: Mark Hammer on October 23, 2013, 12:20:05 PM
If a solid state amp were not desirable or useful for the cleans it produced, I personnally would not bother with trying op-amps beyond the usual 4558.  However, where nice clean highs ARE attainable (and the speakers can support it), trying some bi-FETs or other "better quality" op-amps is not a completely foolish idea.

Yup, putting a set of sockets in there, and experimenting can't hurt. First things first though, going into town tomorrow for caps. Once that's settled, I'll see how it reacts. The first order of business is to eliminate that high pitched oscillation from the second channel. Once it's performing decently, like Paul said, I'll take some time to run it through it's paces, and get my ears attuned to it. Then, I can start monkeying with different transistors, op amps, and whatnot.
  • SUPPORTER
Dead End FX
http://www.deadendfx.com/

Asian Icemen rise again...
http://www.soundclick.com/bands/default.cfm?bandID=903467

"My ears don't distinguish good from great.  It's a blessing, really." EBK

R.G.

I did some redrawing of the power amp into a more othodox style.

See http://www.geofex.com/FX_images/Traced Kustom Power Amp plus mods.pdf

The first diagram filters out some of the obscuring stuff so you can see what's what. I erased one of the two paralleled output stages and the protection circuitry, as well as an oddity about the feedback mechanism that I've never seen in any other amplifier, and just showed feedback from the output of the amp.

The outputs are quasi-complementary, probably to take advantage of the better/tougher nature of NPNs at at the time it was designed. I don't know what the outputs really were, but MJ15003 would be good replacements as would the Toshiba NPNs. The biggest thing to note is that the output stages are simple high current followers, so most modern transistors designed for audio amp work will be fine there. No funny high frequency compensating stuff. The biggest requirement is as big a safe operating area as possible, so you want some of the big modern devices with BVceo of 200-250V and collector currents of 15-20A max.

Modern driver transistors would be OK too. Again, there is a high local feedback, so you're unlikely to get problems with output or output-driver swaps as long as the drivers can produce enough base current.

The second picture shows the voltage gain stage's bootstrapping from the output to give it the ability to drive the outputs fully to the power rails. It gets a little more power out of the same output stage and power supply by driving the outputs closer to the rails, but that's more specmanship than useful range.

The current limiting is pretty standard. The voltage across the emitter resistors (R23, 25, 33, and 35) is coupled to the base of Q6 and Q7 by sensing resistors R21, 22, 27 and 28. When the current across the emitter resistors gets big enough that the base of Q6 or Q7 is turned on, Q6 or Q7 (whichever is handling the emergency...) steals base drive current from the output stage. I think R14 and R15 keep that current limited so output protection doesn't kill either Q6, Q7, or Q3. This is very basic protection, and will not save the outputs if there is a really odd load, but it's good for things like shorted output cables. There are thermal cutouts on the heat sink that take over and will shut things down if the outputs get too hot, but too-inductive loads will still kill them.

The third picture shows some useful mods that won't change the normal operation in any detectable way, but will add some safety factor. The input protector diodes will keep the input trannies alive in some cases. The output catch diodes protect against an inductive load forcing the output beyond a power rail and killing the output device. Normally, they don't conduct at all. The output network is a guess, but I think it will work. The R-C network is fairly standard, as is the damped coil. The coil MUST be big enough copper to carry the full load, which can be upwards of 10A peak. Go bigger wire if you can. #12 Romex is good if you can get it. Wind 20T or so, in layers if you need to, on a 1" to 1.5" coil form - wooden dowel or PVC pipe work well. Slip the wire off the former and solder it in, self-leaded. Mounting these large components may be a challenge.

The bias circuit is an example of very simplistic modern practice. A high-ish gain TO-220 mounted on the heat sink gives good thermal coupling. The resistors make the voltage from collector to emitter be 1+R1/R2 times the base-emitter voltage, and making R2 be a trimpot means you can tweak it in.

The bias circuit is one of those places where disaster lurks. If the bias wires or circuit ever go open, the outputs go saturated on and try to shunt the entire power supply through them. Generally the outputs die protecting the fuses. It's a good idea to calculate R1 and R2 so that if the wiper opens, the maximum voltage across the circuit is about 4 diodes instead of the 3 that would normally be there, or to mount four actual diodes across the driver bases so that if something happens to the wires going out to the sense transistor open, the diodes keep the outputs from going into full melt-down. Also, the wires to the bias circuit on the heat sink carry the full output voltage, so it's good to use covered twisted pair so it doesn't give or receive radiated grak.

There are many more possibilities. For instance, some minor changes to the resistors connected to the protection devices Q6 and Q7 would give you dual-slope V-I limiting instead of simple current limit. The input stage would be a lot cleaner if R6/R7 and C3 were replaced with a real current source. R4 and R5 would be replaced with a current mirror today, it goes on.

But the mods shown are heavy on protection features more than hypothetical performance gains. They should change the actual sound not at all.

I have one of those chassis sitting out in the garage. It's in a head enclosure, but the photo could otherwise been of my amp. I picked it up for $25 at a guitar show, figuring the power transformer and chassis were worth that.
R.G.

In response to the questions in the forum - PCB Layout for Musical Effects is available from The Book Patch. Search "PCB Layout" and it ought to appear.

Mark Hammer

There are some folks in the world whose "spare time" is a precious thing.  :icon_smile:

armdnrdy

Quote from: Mark Hammer on October 23, 2013, 08:59:49 PM
There are some folks in the world whose "spare time" is a precious thing.  :icon_smile:

Says the man with 20,000+ posts!  ;D
I just designed a new fuzz circuit! It almost sounds a little different than the last fifty fuzz circuits I designed! ;)

digi2t

Holy smoly. :icon_eek:

I musta done sumthin' good at some point in my life. That's one heck of a payback if it's the case. :icon_biggrin:

Muchos gracias R.G.! I'll echo Mark's comment, and also offer my infinite appreciation for your thought's and insights on this project.

Going to start digging into it tomorrow, after some cap shopping. :icon_cool:
  • SUPPORTER
Dead End FX
http://www.deadendfx.com/

Asian Icemen rise again...
http://www.soundclick.com/bands/default.cfm?bandID=903467

"My ears don't distinguish good from great.  It's a blessing, really." EBK

Mark Hammer

Quote from: armdnrdy on October 23, 2013, 09:16:11 PM
Quote from: Mark Hammer on October 23, 2013, 08:59:49 PM
There are some folks in the world whose "spare time" is a precious thing.  :icon_smile:

Says the man with 20,000+ posts!  ;D

There's quantity, and there's quality.

digi2t

Darn... Just realized that I only have one of each of the Toshiba drivers. I do however have a pair each of MJ15030/31 on hand.

I know they're not Toshibas, but between the Toshibas, and TIP30/31C's that some people have claimed to use as replacements in these amps, can I run with the MJ drivers?
  • SUPPORTER
Dead End FX
http://www.deadendfx.com/

Asian Icemen rise again...
http://www.soundclick.com/bands/default.cfm?bandID=903467

"My ears don't distinguish good from great.  It's a blessing, really." EBK

armdnrdy

Quote from: Mark Hammer on October 24, 2013, 09:30:32 AM
Quote from: armdnrdy on October 23, 2013, 09:16:11 PM
Quote from: Mark Hammer on October 23, 2013, 08:59:49 PM
There are some folks in the world whose "spare time" is a precious thing.  :icon_smile:

Says the man with 20,000+ posts!  ;D

There's quantity, and there's quality.

I can attest to the quality of your posts! You have helped me and others immensely!
If fact....you have helped me at times in spite of myself!  :icon_wink:

Dino,
If you can't find a replacement for the Toshiba drivers....and don't mind waiting a bit....try Utsource.
Also, try Consolidated Electronics. Consolidated has quite a large stock of Japanese semiconductors.
I just designed a new fuzz circuit! It almost sounds a little different than the last fifty fuzz circuits I designed! ;)

R.G.

#39
Quote from: digi2t on October 24, 2013, 10:23:45 AM
Darn... Just realized that I only have one of each of the Toshiba drivers. I do however have a pair each of MJ15030/31 on hand.
I know they're not Toshibas, but between the Toshibas, and TIP30/31C's that some people have claimed to use as replacements in these amps, can I run with the MJ drivers?
I think the MJ15030/31 would be fine drivers for this amp.

One question - are the original drivers OK, or blown? The biggest stress by far is on the output devices, not the drivers.

A TIP30 or TIP31 would make a decent Vbe multiplier for bias.

Amp repair dirty trick: when you replace the outputs and/or drivers, the bias network is suspect until you get it tweaked in. Put a shorting wire across the bias diodes and/or network until you verify that the outputs and drivers are working right. It will have horrible crossover distortion, but you'll know that the bias setting is not messing you up. When you verify outputs and drivers are working OK, you can remove the short and adjust the bias properly.

Also - a light bulb limiter will help you ensure not frying things while messing with the output stages.
R.G.

In response to the questions in the forum - PCB Layout for Musical Effects is available from The Book Patch. Search "PCB Layout" and it ought to appear.