Need amp repair help - Can you ID this semiconductor?

[GaryM]

Greetings!

I need assistance in identifying a semiconductor. It is from a ROAD 220 bass guitar amplifier. The part was mounted on the driver pair heatsink and, in circuit, it is wired across the PNP/NPN driver compliment from base to base. I'm really stumped because of the odd ID, and odd case style. I'm sure it is part of the thermal regulation of the output xistor bias. I'm unsure if it is either a transistor or a diode array.

Google has not turned up either a schematic for the amp or anything that remotely matches the markings on the part. One leg of the device is completely blow off. The drivers and output transistors were all shorted C-E.

Thank you for your help!

Gary

[numpsha]

  No real help here, but did you check schems for Rickenbaker Road series amps? -- unsure how much of the original circuitry they retained after buying out Road amps, tho. Location and damage sounds like diodes. What did third leg tie to?
  BTW, I'd be glad to take that old hunk of iron off your hands to save you further grief. (wink, wink)

rob
 

[GaryM]

Thank you for your offer to relieve me of my grief.  I don't know if it's an NPN, PNP or what the pinout might be. As you can see, the device is literally blown. I've searched the Rickenbacker stuff and there was nothing for my amp. I was hoping someone might recognize it. It looks like this head...

[teemuk]

I've seen the picture before. (Honestly, is this the only pic of this amp in Internet). It's a Road 220 Bass. Reputedly very, very good amplifier built solidly. Had extremely efficient loudspeaker cabinets too. Unfortunately they're pretty much unknown today. Trying to get schematics to Kustom amps (another company of the man behind Road amps) is very hard – I believe that finding schematics to these things is nearly impossible. Anyway, I took a liberty of sending Rickenbacker email on your behalf (I'm curious about Road schematics as well) and enquired whether they have any schematics to pre-Rickenbacker era products (or any Road amp schematic close enough to Bass 220). Most of you don't likely know it but they only share the "smallest" schematics at their website. I haven't heard from them yet but I'll let you know what they answer.

Judging by your description at diyaudio.com the overall circuit sounds like the basic 3-stage Lin topology with quasi-comp output. Similar to what most SS power amps were since the early 70's. The device is very likely a NPN transistor (although if both resistors from B to C and E are the same value I think a PNP would work just as well). The sub circuit, in which the transistor is, is called a bias servo or "VBE multiplier". You can find a lot of discussion about it from the Internet. Heck, you can even download my book to read a lot of stuff about it that is not covered so often. Basically the ratio of the resistors determines a certain voltage drop over the transistor, the voltage drop consequently biases the output devices. I think you can substitute the transistor with any model for experiment but heed that the beta value drastically affects the circuit's operation – so does the thermal coefficient. There's a good chance that almost any transistor with moderately low-beta will work sufficiently, though. (It's a semi-power device so I assume the beta value is not likely very high). You can monitor the bias and temperature and if they seem to go way off the servo is not working correctly. If the amp has a lot of crossover distortion it means the servo cannot develop a high enough voltage difference and you need to alter the ratio of the resistors. If it doesn't have a bias trimmer then one might be a good addition (however, there are certain rules for it's optimum placement in the circuit). You don't need to replicate the old circuit exactly to get it working, any modern device will work perfectly if the bias circuit is tweaked accordingly.

Aside replacing the blown semiconductors I'm pretty sure it will require a good overhaul that involves potentiometer cleaning and replacing all aged electrolytic capacitors. I would love to see more pics of this thing.

[GaryM]

Teemuk, if you sent them an email on my behalf, I must return the favor. Here are some pictures for you to check out...

The first is the power amp board which is out of the amp. I have replaced all the defective components and you can see where the unknown part mounts. (See the missing leg?) There are no bias pot or diodes in that portion of the circuit.



Upper left is the PNP input differential pair. Down and slightly right is a TO-220 NPN, class A driver, followed by the complimentary pair of drivers on the heatsink that I replaced.

Being as there are no other photos on the net, here are a few more for everyone to check out, sans the amp head cabinet. I'll post those after it's operational and reassembled.





It deserves to be brought back to life.

FWIW- I sent the semiconductor picture at the top to Technical Support at NTE. They replied with, "Just tell us the part number and we can sub it for you." 

Gary

[GaryM]

I broke down and sketched the driver board. You can understand my dilemma with this. - Gary

[teemuk]

Haven't heard anything from Rickenbacker yet but I guess I can try to help anyway.

I don't know how many of the transistors were toasted besides the one in the VBE multiplier but TIP31C is pretty close replacement to NTE291. (C is essential because it marks the highest voltage rating) I suppose you can therefore sub the drivers with TIP31C and TIP32C. VAS can be TIP31C as well. As was mentioned before, the VBE multiplier transistor is something you have to experiment with. I would try some garden-variety transistor there (i.e. BD139/BD179); likely you don't even have to alter the bias circuit in any way. The thermal stability resulting from tracking only the driver temperature is pretty questionable, though.

Aside the strange 220-ohm resistor in series with the bootstrap capacitor and the 220-ohm resistor at the VAS collector the circuit looks very typical. Yeah, I wonder who put the 1K-resistor there – that's been a total hack job!

The things that worry me are the 220- and 470-ohm rail resistors. I think they should be a decade lower value, at least the 220-ohm one should be. Check your readings or see if they look replaced. If I have figured this out right the 220-ohm resistor biases the PNP driver incorrectly, which creates enormous power losses over the "PNP" side output transistor. Even at idle. The 470-ohm resistor in turn drops the voltage too much and consequently restricts the voltage swing of the voltage amplifier stage making it clip too early. (I don't know the rail voltage but I assume it is about ± 40V)

I tested the circuit with SPICE using rail voltage of ± 40V, the aforementioned resistor value corrections and transistor substitutes and the simulator showed the circuit putting out 220W peak into four ohms, which - I suppose - is how the circuit should perform.


By the way, what is the output transistor type and were there originally three pairs of them? (I counted the mounting holes). That would explain the additional current sharing resistor at "PNP" side...

[tonyharker]

Apparently ballasting the bootstrap capacitor with a resistor helps prevent turning on fully the NPN Driver transistor with large positive-going signals, which could cause this transistor to 'lock' hard on.  This of course would subsequently destroy the output transistors. 

Tony

[GaryM]

By the way, what is the output transistor type and were there originally three pairs of them? (I counted the mounting holes). That would explain the additional current sharing resistor at "PNP" side...

There were only two, NPN, TO-3, transistors mounted. (RCA transistors with a house p/n.) The extra mounting holes were not used. (Yes, I've already been tempted to hang more output transistors there and install a higher current power transformer.)

I'll go back and redraw the schematic after I get this thing running. I likely erred in the 220 ohm resistor in the negative rail. It looks like that 1K resistor in the output collector should be ~0.33 ohms as that is the take off point for the current limiting circuit for that side. I should drop that 470 ohm 1 watt resistor down to ~100 ohms, too.... and add more filtering.

It would be best to mount the thermal regulating device on the output transistor's heatsink, as opposed to the driver heatsink, which just adds another thermal junction.

So which do you think would work best?

A) Use and NPN device with the emitter connected to the PNP driver's base and the collector to the NPN driver's base.

or

B) Use and PNP device with the emitter connected to the NPN driver's base and the collector to the PNP driver's base.

Thanks for helping me hash this out. Talking it over sure makes the troubleshooting easier.

[GaryM]

Finished a major redrawing of the power amp stage. Here is an updated schematic that includes the current limiter circuit and corrections to my hand drawn sketch. I'm sure a few of the odd resistor values are due to a prior repair. It looks like some circuit value changes are in order to balance things out better.

Gary

[teemuk]

I see that the 220-ohm rail resistor has swapped place. Intentional?

Shouldn't there be a voltage divider resistor from base to feedback loop at PNP side current limiter as well?

The VBE transistor: I would try an ordinary PNP first. If it doesn't work sufficiently I would likely add a trimmer pot. The thermal sensing of the drivers doesn't "go by the book" but I wouldn't change it. - Or at least that would be lesser priorities and only done if the thermal stability wouldn't seem to be in order.

Any Zener diode in the range of 10 – 15 V works fine.

I'm inclined to think that only a single pair of power transistors is not enough and maybe the previous repair guy just removed the rest. If the amp produces an output power of 220W peak it means that the output devices are dissipating a total peak power of more than 100 watts. Are they rated sufficiently for this? Even if they are the extra pairs would do no harm.

I still haven't heard anything from Rickenbacker.

[GaryM]

Thanks, teemuk. I see what you mean and will modify the limiter circuit before firing it up tomorrow. There is no hole on the PCB for a resistor for the PNP limiter from base to emitter.

Yes, I had that 220 resistor in the negative supply line on the wrong side in my hand drawing.

Regarding  the pair of TO3 devices, there is no "original" wiring for more. All the associated wiring and connectors are soldered to the driver board and there are no attachment points, clipped wired or broken solder connections to indicate otherwise.

You're gonna get a kick out of seeing the massive heatsink the amp has. It gets attached to the chassis and not to the output transistors!

More soon,

Gary

[teemuk]

I've seen few photos of a 220 "Lead" model: The heatsink arrangement is quite unordinary, and yes the heatsink is massive. I also found out that Road Electronics introduced a 440 model as well so I take back what I said about those transistors. (I would still use all of them, though).

[tapokc]

I have a similiar Road guitar amp with the same board in it.  The part in the same location on my board is a 2n6474.

Thanks for the info, I can really use it with my amp.

Did you have any problems with any of the pots, I have two gone and I need to replace them.

Thanks and good luck.

[GaryM]

You are welcome! The 2N6474 on my board has a TO-220 case. The bias device was a different case style altogether. I've set this project aside for another day as it's been difficult for me.