• TalkBass has been independent since 1998. Add your voice.
    Create a free account to reply to discussions, view embedded media, and browse with fewer display ads.
    Join free Log in
    Want zero display ads or expanded classifieds tools? Compare plans.

BVV3000 INFINIUM Trial

#161
seamonkey said:
There are some absolutely amazing electronics tear down videos and tutorials showing some have figured out amazing complicated circuits. It's fun to see how much people can figure out, including hacking the code as I'm sure there's a CPU in the Infinium.
Click to expand...
Yeah, I already traced which connector senses the current across the 4.7 ohm cathode resistors, and which connector feeds the negative bias voltage. The other connector with a decent amount of wires goes to the indicator led's, just have to figure out the 4-5 other plugs off the infinium which are all only 4 pins/wires or less.

The plate has an inductor in series, the screen uses a 22 ohm.
I printed out some zoomed in sections of an svt to use as my notepad.
 
#162
The one thing I learned and can recommend to anyone working on one is to make sure and test it with a known good working full set of output tubes!

I put in a full set of good tubes this morning and got completely different results than before.

Every tube autobiased to about 35ma at 600 b+. These tubes are not matched well at all and the bias voltage was completely different for each to make the current the same, as it should. I measure about .164volts across each 4.7 ohm cathode resistor.

I think the amp doesn't like missing tubes in sockets. I am also assuming that it auto balances the sides so when it was missing tubes it compensated by upping the current on certain ones to balance to the other side.

I'm going to do a lot more testing and I still need to check the bias of the tubes that came with it. 5 of them tested good on my Hickok but I didn't put them in another amp yet to check how they bias.

It's interesting to troubleshoot since the amp keeps autobiasing itself and seems to turn on initially at a cold safe setting and then after about a minute it starts calculating.
 
  • Like
Reactions: seamonkey
#165
As you have already stated, the infinium circuit continually monitors and adjusts the bias from what I can see. After the first minute or so the LEDs flash randomly in no particular pattern that I can recognise, sometimes all off for a while then a couple flash then all sequence in sweep. Each LED only stays on for a fraction of a second though.

Here's a link to an interesting article by PrimaLuna amps which also claims evidence of greatly increased tube life due to a similar(?) auto bias technology

Adaptive AutoBias
 
#166
So I did actually find a problem with the amp which perfectly explained why it seemed to sometimes work and sometimes not.

I was testing it and measuring voltages and noticed when I bumped a wire the infinium lights would flash. This was repeatable and when I tapped them enough the lights would blink in a sequence like it was resetting and the bias would be all over the place.

This cable is the one that sends the cathode voltage across the 4.7 ohm resistor to the infinium board. I removed the cable and connectors and hard wired it.

The amp works exactly as it should and I've swapped several sets of power tubes in and it always biases them to about 35ma within a couple minutes.

I checked output power before clipping and this is not a typical svt like 300 watts. The most I got was just under 200 clean @ 250hz. At 100hz it was closer to 190 or less. Any 6550 svt that I have tested when setup right does about 270 clean.

This only has 565 volts b+ vs 660 of an svt so the numbers make sense. It's also biased a little cool but not too bad. I only saw crossover distortion on the set of tubes which were furthest away from matching each other.

The circuit boards are made pretty good. I definitely had to use some heat to remove the connectors and I didn't lift any pads or traces. As a comparison, I've lifted some on boogies much easier.

I've only bench tested it so far and need to hook it to my cabinets and check it out.

I wonder if you could get a power trans from a new svt to boost the power a bit. I think it would work because the main trans on both only supply the high positive voltages. Everything else comes from the filament trans.

The other five tubes were fine, just the one bad one when it came to me.
 
  • Like
Reactions: seamonkey, middy, Lindex and 1 other person
#169
The XFMR idea is nice. Are there any Bugera components downstream of the transformer and are they rated high enough to take it?

BSEE for 25 years here. Did tech school before college to learn how to maintain my SVT(in the 70's) and hot rod Fenders and Marshalls for guitarists before factories used master volumes and such.
 
#174
I used:
a set of JJ KT88's, Sovtek KT88's, Sovtek 6550WE's, Peavey (china) 6550's and got the same results +/- a few watts. The worst looking signal was from the sovtek 6550's but those tubes were the worst matched set I have.
In a 'normal' amp they would bias about 10-15 ma away from each other so those tubes showed signs of crossover distortion.
 
#178
The amp? It's already working. Or modding to try and get higher bias current.
I think it is at the most power it will put out. Need more B+ voltage to push the tubes harder.
I suppose trying to push to 45ma might be worth a shot. Just to see.
Need to get some 1% 3.7-3.8 ohm resistors though.
 
#179
I was asking about the amp. I thought the last thing I'd read was that the problem had been identified & resolved. From there I thought there still needed to be tested with cabs attached. Maybe I missed something.

Regardless, I have enjoyed reading about your efforts to get it going. So much so that I have an increased interest in all-tube heads. Nice job and thanks!
 

Latest posts

Top Bottom
Back