• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Quicksilver 8417 bias runaway?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Dear Quicksilver & VTL users,

You're gonna keep blowing up tubes, no matter what tube you put into those amps, or VTL amps for that matter (very similar, eh?)!:xeye:

Why?

Simple, they are in a "rating war" for maximum output power. Do you really need 1.0 to 1.5 dB more power than the tubes are rated for? Will you hear it? Well, it does make for better advertising specs.

The problem is (assuming no leaky grid coupling caps) that the screens are being run way way too hard, on peaks they are substantially beyond design maximum, even for NOS USA vintage tubes!

(typically the B+ is also near the hairy edge of reality too...)

I've repaired a number of these amps, ALL have the same problem. The tubes have short life, and tend to run away and some (literally) burst into flames.

To fix, reduce the maximum screen dissapation so that at full power into load the screen current is NOT over the actual max for the tubes you have in the amp. Usually that is a single resistor in series with the screens.

Hint: if you put a dummy load on the amp, run it to full power and look carefully at the inner structure of the tube (you can on most typical output tubes) the screen should not be glowing white like a lightbulb filament. ;)

That will keep them from blowing up and having short lives.

It will ALSO reduce the power back to the normal rated power for the tubes.

My advice is to get over the "loss" of power. If you really need more power, then you have the wrong amp.

Consider building your own amps, seriously.
These old circuit designed amps are very lacking in so many ways compared to what is possible from tube amps that it isn't funny.

At least adjust the screen current into range.

_-_-bear
 
The Sylvania tube manual lists a maximum grid circuit resistance of 47K for a 8417 operated in fixed bias. The Quicksilver uses 150K, presumably to make it easier for the Cathodyne phase splitter. Me thinks dropping the value of the grid circuit resistance would lessen the biasing problem, but you would probably want to increase the idle current for the phase splitter section.
 
Burnedfingers,

Suggest measuring the current through the screen resistors for good luck. I think they made the new "KT-88s" pretty overbuilt, from what I am told. Dunno. Measure the current at full load (also what freq??).

Arnoldc,

Perhaps that is why. A clue to be sure.
Don't overlook the potential for parasitic oscillations.
Again, the easiest thing is to just measure the current through the screen resistors - or if there is none there, put a small one there and measure the current.

_-_-bear
 
Quicksilver 8417 Tube Options

Hello, I'm new to this forum and I have followed this discussion from the beginning. I also have a pair of Quicksilver 8417s. Recently one of my amps which currently has one RCA and one GE 8417 tube has been getting out of bias. I talked to Mr. Sanders of Quicksilver on the telephone yesterday and he said that 1) ideally I should use tubes from the same manufacturer on the amp 2) I should set the bias at 90 when cold and 3) GE 8417 tubes tend to run away as stated by others in this thread.

Now for my question, I have an opportunity to buy a couple of new old stock Westinghouse 8417s. Does anyone know if these are good tubes like the original Philips or Sylvania or are they going to be like the GEs ?

John
 
I highly suggest that only Sylvania tubes be used. I had a learning curve with different brands of 8417 tubes and found the Sylvania to be less likely to run away. I have a 3lb coffee can of 8417 tubes that are going to stay right in the can. The amp sounds real close to the 8417's using KT88's So I have been using KT88's for years.

In order for the 8417's to work without run away they have to be matched perfectly and as you asked the grid resistors need to be changed to make the amplifier more stable. Another modification would be the bias circuit as individual tube bias is necessary with the 8417's to make sure one doesn't flame out.

As Mr. Sanders mentioned that you set the bias LOW when cold. Do not walk away from the SOB because the minute you do there will be trouble in River City. After the amplifier is warm 30-60 minutes then run the bias to about 100mA for the pair.

Do not get talked into an EL34 replacement because they totally suck and the price is too much for the small abount of work done in my opinion.

Did you just get these amps off ebay or were you trying to sell them there recently? I saw a pair with mismatched output tubes recently.
 
I have four NOS 8417s, all Sylvania-manufactured. The labels on three of them say "International Servicemaster" (and they're in boxes that say International Servicemaster), and the fourth says General Electric (and is in a replacement GE box, as the original was destroyed by moisture in storage)... but ALL FOUR of them have the CHARACTERISTIC LABEL scheme on them that ONLY Sylvania used- looks like this:

8417
USA

(in black ink, on the side of the tube)

BTW: These are GENUINE 8417s, and not "fraudulent relabels", as some have apparently found with some 8417s from "jobber vendors" (such as Int. Servicemaster or IEC or such). I have verified that they have the PROPER SIZE AND SHAPE plates, for Sylvania 8417s. That, along with the factory-printed Sylvania type code (as described above) definitely settles it as correct.

One note- all four test as new on the Sencore Mighty-Mite (full emissions, and no shorts). I don't have an 8417 setting chart for the Hickok TV-7A/U here, so I can't test them in there... if anyone has the settings for such, I'd love to run them through the Hickok...

Jekman- Email me if you like ( gwaters@inbox.com ). I don't have anything to use these in now, so I can stand to sell them, if you need them...

Regards,
Gordon.
 
Burnedfingers:

Thank you for your valuable advice. I hope to replace the two mis-matched tubes ASAP and monitor the bias closely to see how it goes.

Yes I was the buyer on the ebay auction ended 12/30/07 for the pair. I wanted the Quicksilvers because the company is local, they have a great sound and they would be a good match with my Tannoys. From the Ebay description, I knew that the tubes on one amp were mis-matched and might be trouble. The seller said that he listened to both amps and they sounded fine, which meant that they at least worked! Besides, compared to buying McIntosh MC 75 s, they were less than half the price.

Anyway, once the auction ended I got busy finding matched pairs of 8417 tubes. I bought two matched pairs of Bogens by Sylvania (I am waiting to receive them). I also wanted to find another couple as back ups which is why I was looking at the Westinghouse tubes. It sounds from your experience that I should just switch to KT88s. Although would'nt that require some modifications?

Regards, jekman
 
I have four NOS 8417s, all Sylvania-manufactured. The labels on three of them say "International Servicemaster" (and they're in boxes that say International Servicemaster), and the fourth says General Electric (and is in a replacement GE box, as the original was destroyed by moisture in storage)... but ALL FOUR of them have the CHARACTERISTIC LABEL scheme on them that ONLY Sylvania used- looks like this:

Are these the ones you were trying to sell for $225? That a tad steep isn't it?

I put a new matched quad of Sovtek KT88's in mine for less than $100 shipped.
 
burnedfingers said:


Are these the ones you were trying to sell for $225? That a tad steep isn't it?

I put a new matched quad of Sovtek KT88's in mine for less than $100 shipped.

It's as low or lower than what I saw 8417s listed for anywhere else. :confused: Try to buy them from any of the tube vendors. Many are selling on Ebay for $75 each (I just looked at completed auctions). The GE tubes can be cheap, but the Sylvanias are usually at least $65 each or so (with rare exception, and it looks like those cases, the people are "making up" the difference with shipping "costs")...

Also, you gotta factor in the cost of having the amp modified... and truthfully, a KT88 is never going to quite sound like an 8417. Just doesn't have the transconductance (like 23000 for the 8417, compared to 11000 for the KT88/6550, in the same conditions!) . This leads to a lot higher output impedance for the tube (like twice the value of the 8417)... you'd have to use a lot more negative feedback to get the same output impedance at the output of the amp...

That's the "dirty little secret" about the KT88 mod... yes, it works... but it's not the same amp anymore...

Also- as for INDIVIDUAL BIAS- that's the ULTIMATE SOLUTION. This is the same situation that people are having with amps such as the Futtermans and such, which use now-rare tubes like 6LF6s and such... the smart people are changing them over to INDIVIDUAL TUBE BIAS, so that having to buy matched tube sets (for bias current/grid voltage) IS NOT NECESSARY anymore. On top of that, many times, the amps sound BETTER than before (because you can match the bias MUCH MORE PRECISELY with a pot on each tube, than ANY set of "matched" tubes will match on their own!). If you're going to modify the amp, this is the SMART way to do it!

Regards,
Gordon.
 
Also, you gotta factor in the cost of having the amp modified... and truthfully, a KT88 is never going to quite sound like an 8417. Just doesn't have the transconductance (like 23000 for the 8417, compared to 11000 for the KT88/6550, in the same conditions!) . This leads to a lot higher output impedance for the tube (like twice the value of the 8417)... you'd have to use a lot more negative feedback to get the same output impedance at the output of the amp...

Factor the cost of being modified? This is DIY land my friend. The land where we do what others pay to have done. Anybody can successfully modify a 8417 mono block and have it turn out ok.

Have you modified any 8417 amps and listened to them?

I have

I have owned 4 set of 8417's and one set of KT88 mono block amps.

Side by side there is VERY little difference between a 8417 and a
8417 that has been modified. I have had people do a listening test. Slight difference yes....worth $75 a piece tubes? Not to me it isn't.

Yes, I played with feedback via an adjustable pot I installed.
 
burnedfingers said:
k.

Have you modified any 8417 amps and listened to them?

I have

I have owned 4 set of 8417's and one set of KT88 mono block amps.

Side by side there is VERY little difference between a 8417 and a
8417 that has been modified. I have had people do a listening test. Slight difference yes....worth $75 a piece tubes? Not to me it isn't.

Yes, I played with feedback via an adjustable pot I installed.


Funny you should mention that.

I work, as I have for many years, in a speaker repair shop that also does audio repair. One of our regulars has a pair of these amps.

He actually DID the mod, not long ago... in fact, he had us order the quad of KT88s (Ruby Tubes, made by Sovtek, but more carefully tested than most other vendors) for him, and the replacement caps and bias resistors to do the mod, as described on the Triode Electronics page.

Once he did it, he called me... and mentioned that the bass just wasn't as "tight" as before, on his Vandersteen speakers. I went and looked up the data, and we had a conversation about transconductance vs. effective output impedance. He had done the phase splitter mod (paralleled the sections of the 12BH7), but still couldn't get it back to the way he liked it.

He read, somewhere, about possibly changing the feedback (I think on Audioholics, or somewhere like that), and tried that... changed the feedback feed resistor and the resistor on the input tubes, several times. Only thing, is that he complained about having to turn up the volume control higher on his preamp, significantly, afterwards, compared to before the mod... accompanied by a "loss of dynamic WHAM"... I think that was the phrase he used. THink he may have been getting too close to clipping on his preamp... that was the only thing I could think of, that would cause this...

He finally found a resolution- our tech here suggested changing out the phase splitter tubes for 6201s (high-grade 12AT7, higher gain AND transconductance than 12BH7s or 12AU7s). They got the phase splitter re-balanced with the new tube (changed out cathode and anode resistors, and IIRC, may have even tapped a higher B+ from somewhere in the power supply to run it- can't remember exactly)... and THEN there was enough gain. Had to play around with some caps, IIRC, to get it to be happy (stability?), but I do not remember the exact specifics, unfortunately.

Only after ALL THAT, was the amp acceptable, as a close approximation of the original for him.

Sounds like pretty much re-designing the amp, to me. Glad he had time to go through all that... it took several weeks to figure out all this stuff...

Naturally, some people's mileage may vary... if you've got a preamp with infinite output voltage, then you may not have this issue, to the degree he did... but it was an object lesson to me...

Regards,
Gordon.
 
Did you hear this mod yourself?

Most people don't bother paralleling the 2 sections of each tube together. Quicksilver only uses one section of the 12ax7 and one of the 12bh7 in stock form.

Sorry Gordon been there done it. Listened to a bone stock 8417 next to a convered 8417 and they both hit hard. There is however a slight difference in detail between the two but then again the KT88 tubes were new and straight out of the box. But then thats my opinion based on doing a few of them and owning a few of them. I'll take a slight decrease in performance to having the tubes turn red and taking out the pair of 5AR4's. If I really thought the 8417's were that good I would be using the dozen plus tubes I have.
 
I enjoy this tube in place of the 8417 :
 

Attachments

  • toshiba-6gb8.jpg
    toshiba-6gb8.jpg
    91.2 KB · Views: 209
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.