How much ripple is too much?

[Chuck H]

Here's a post I found on the Kustom forum (when there still was one). I actually had to do a little further research to find a couple of obsolete preamp parts that had cascaded four different designations over the years and the last incarnation still had to bebought NOS for too much $$$. But I hope this helps. This is just as I copied it:

ua739, RC4739, XR4739
Q1 Q2 NTE159 / Q3 NTE190.
Q4 & Q8 are NTE128, Q6 2N4401,
Q7 Q5 Q9 are 2N4033
Q10 thru Q13 are 2N3055.
CR3 and CR4 with 1N5402,
CR1 / CR2 / CR5 are the originals but test good.
Here’s something that might help. There are several oddball transistors in the amp. I just finished repairing one today. There are two driver transistors on the power amp pcb, 38736 and 38737. Well guess what… NTE 128 and 129 work like a charm (note: there is no listing in the NTE book) and solder straight in. One small drawback: the original transistors are welded to the heatsinks. What you have to do is drill them out and place the heatsinks over the new transistors. The 2N4249’s and 2N3638’s are in the NTE book (NTE 159 and NTE 129, respectively). There are a buttload of transistors marked S35677 (that do not cross). Well guess what? A 2N2222 works like a champ. The price for a 2N2222 is 30 cents per pop. I found one bad one (totally open) on the large board and one that was borderline weak on the small board. I didn’t have to replace any output transistors, but my guess is that a good ole 2N3055 should work. That is a general purpose NPN power transistor. The other thing that sucks is that the circuit boards are attached to the front panel via the pot shafts. Here’s the other goofy thing- the pots are not your usual PC mount controls so one would have to do some serious jury-rigging (a Navy Firecontrolman has this sort of thing down to an art) to put in a replacement pot. In a lot of SS amps, NTE 128 and 129 work like a charm for driver transistors.

[AMPREPAIR]

Thanks for providing the equivalents! The transistor information you provided seems to match-up nicely. I just hate to buy overpriced replacement parts from NTE.

[soundguruman]

Yeah, check ALL the solid state components, if the drivers are shorted, there's bound to be MORE.
What makes you think the outputs are not shorted?

[Chuck H]

When I had one of those in I had to work from this list and continue googling for other cross reference numbers. It's was a long tedious chore.

[52 Bill]

When I had one of those in I had to work from this list and continue googling for other cross reference numbers. It's was a long tedious chore.

The driver transistors were RCA 40409 and 40410 transistors with house numbers for Kustom. They are on eBay at times, NOS. I know that there are any number of suitable replacements for these as long as you don't need the heatsink to complete a circuit path, as they are directly connected to the collector and would be used as a jumper to complete circuit paths.

The basic output amp is so common a design, you can use almost any transistor in there and get good results. Try using the transistors that Peavey uses for drivers or even TO-220 TIP's. I think that the originals were rated at 90 volts and 1 amp.

And the Vintage Kustom site has a basic cross reference posted there. You'll need to search it out.

[Chuck H]

The hard to find amps are on the preamp boards. And the post above is a pretty good start for cross reference since most of the Kustom amps use the same parts. Since the amps are built on fairly crude PCB's it's best to confirm pinouts. Winging it on a PCB with the wrong pinout is a PITA.

And for the record, Bill, find me one of these if you can. Note that it has no KPN (Kustom part number). I think I may have found an ancient data sheet for it and one forum reference that only theorised on how to replace it with a MOSFET but no cross reference numbers.

EDIT: SOB... It's NTE221 cross reference. But I swear to Pete this info couldn't be found only a few months ago?!?

[52 Bill]

And for the record, Bill, find me one of these if you can. Note that it has no KPN (Kustom part number). I think I may have found an ancient data sheet for it and one forum reference that only theorised on how to replace it with a MOSFET but no cross reference numbers.

I probably have one on an old board around here if you really need one, I can't remember ever replacing one. The generic number is RCA 40841. It is a dual gate mosfet made by RCA. NTE lists NTE221 as a replacement. According to the schematic that you posted, the Kustom part number is 007-0024-00. A quick Google search shows that Parts Express has the NTE221 in stock for $15.

Dual gate mosfets were common back in the 70s and were used in a lot of radio circuits. There are other numbers that would probably work as well.

[R.G.]

Thanks for providing the equivalents! The transistor information you provided seems to match-up nicely. I just hate to buy overpriced replacement parts from NTE.

That's smart. However, you can also do what NTE does - get things that are equal to or better than the original. Back when the original Kustom amps were being designed, it was hard to find power transistors with enough voltage, current, power (and that's different), and SOA (different still) to run an audio amp, and they were expensive. They were $4-8 in 1970s dollars, which is the equivalent of $12-24 dollars today. (As a curious side note, no government in recorded history has been willing to resist the temptation to inflate their currency if they had a chance.)

Today, you can get much better power and signal transistors for a pittance. Toshiba, Sanken, and On Semi make devices which will exceed the specs of those original devices in any way; the only tricky thing is that the metal TO-3/TO-204 case is hard to find and expensive. If you can live with putting TO-247 flat packs onto the heat sinks, you can get better-specs devices for under $3.00 in today's micro-dollars.

Same for drivers. The 40409 and 40410 were ground breaking at the time, but you can get better-in-all-specs drivers for well under $1.00 each in today's micro-dollars in TO-220. And signal transistors are almost free.

If you insist on TO-3 power devices, the MJ15024/15025 complementary pairs are good as are all the MJ150xx series. If you can, get them from a well-known distributor because they, like Toshiba's power outputs are so commonly counterfeited that I would not trust anything but an authorized dealer. Newark says they have MJ15024s and 25s from On Semi in stock at about $5.00 each. MJE15034/35s from On Semi are in stock at about $1.35 each. If you can use plastic devices, the NJL1302/3281 On Semi devices are in stock at about $3.25 each. All of these are vastly more capable power devices than the originals.

[R.G.]

For those who aren't familiar with these amps... The filter caps look like a pair of coffee mugs with screw terminals on top. Mammoth! The ones in the amp I worked on were a different brand from stock, so I assumed they had been replaced. But I priced similar caps in case it came to replacing them. VERY expensive. Prohibitively so.

The old "computer grade" caps. Very expensive today. And also passed by by technology.

Today's caps are much more compact for the same voltage*capacitance rating, as well as being rated for more ripple current and lower ESR. They just come in snap-in PCB mount packages. If you're disciplined and creative, you can replace the old capacitor-megaliths with arrays of PCB mounted snap-in caps. This is much less complex than getting custom PCBs: you figure out the footprint of the snap ins, and lay out an array of holes for the leads on a sheet of copper-clad. Drill all the holes, clean the PCB, and paint the copper lines for all the +'s, -'s and grounds, then throw it in the etch tank.

...er, also having pre-drilled the mounting holes first...

You'll get as many uF and V, but lower ESR, higher ripple rating, and a lower operating temp by spreading out the ripple current heating over a bigger area. And probably cheaper.

I recently bought two dozen 18,000uF/63V snap-ins surplus for $1.00 each. They're 1.375" diameter by 2" long.

[J M Fahey]

Did you check for DC across the speaker out terminals?
How much do you have?
With the speakers connected and disconnected.
Thanks.

[R.G.]

Excellent advice, JM! An amplifier cannot be working right if it has more than about 10-50mV on its output. 0mV offset is desirable, but some millivolts is tolerable as an offset. More than that is not, and indicates a flaw.

[soundguruman]

Excellent advice, JM! An amplifier cannot be working right if it has more than about 10-50mV on its output. 0mV offset is desirable, but some millivolts is tolerable as an offset. More than that is not, and indicates a flaw.

On a Kustom, there should be NO DC on the output, none at all.
If there is DC, you at least have mismatched output transistors, and that "does" cause hum in these amps.
The amp is not tolerant of mismatched outputs.

[AMPREPAIR]

Fortunately, there is "0.0" VDC on the output (with and without load).
I am still working on cross referencing the transistors (time consuming). A silk screened PCB would make this repair a snap!
The hum remains when the power amp input jack (J1) removed - so it seems everything before the power amp is eliminated as a possible hum source.
Someone asked about the output transistors. They are new and check fine.
I would still like to know a "rule of thumb" that may be used to determine how much ripple is too much?
I am frantically working on "too-many-repairs" and it takes a bit of time for me to reply......sorry.

[Jazz P Bass]

Rule of thumb?
It really does depend on the power supply.
Anyway, on yours (+ - 40Vdc) I would thumb it as anything over 200mv's will cause a problem.
Another rule is 'how close is the Vdc of the + & - rails' when compared to each other.
By the way, have you checked the low voltage power supplies.
In the opamp supplies especially, you do not want to see any Vac ripple.

[g1]

I seem to recall a figure of 5% used as a rough rule of thumb.
Hopefully someone will correct me if that number is out of whack.