kt88's in a 2203
Moderator: VelvetGeorge
-
- Senior Member
- Posts: 386
- Joined: Fri May 12, 2006 11:34 pm
- Location: Oahu HI
kt88's in a 2203
I'm going to throw some kt88's into my 2203, is there any mods that I would need to do besides just biasing?
Children of the Grave!
- wdelaney72
- Senior Member
- Posts: 1619
- Joined: Sun Jan 22, 2006 9:04 pm
- Location: Chicago Suburbs
Here was an explanation I got for 6550 v. EL34. KT88 tubes would be consistent with 6550. I apologize, as I can't recall who sent this to me, but it was one of the veterans on the forum here...
The bias circuit, for EL34s (B+ around 450V or so depending on wall voltage and bias) it should be 220k in series with the diode in the bias circuit and 56k in line with the bias pot, plus 220k for the grid load/bias return resistors. For 6550s 150k, 47k, 150ks respectively. The R in series with the bias diode gives more negative voltage if you go lower, the R in series with the bias pot gives less as you go lower, and the grid load/bias return Rs won't effect the bias voltage but can have influence on the stabilty of the outputs. The R in series with the bias pot and the bias pot itself make up a resistance from C- (the bias voltage) to ground--the less the resistance is the less negative voltage it gives. The fixed R (47k or 56k) in series with the bias pot makes sure the adjustment range is restricted and you don't short out the C- to ground running the tubes too hot and damage them. The R in series with the diode looks kind of odd maybe brown red (sort of looks orange but maybe it's actually yellow?). If it was 120k, I think could mean that it has a lower B+ PT as found in JMP 50Ws so that was the case 120k would make sense (bit lower B+, bit lower negative bias voltage range). If you want to know what the resistances are, the best thing is to just measure them (amp unplugged, voltage drained, etc.).
The bias circuit, for EL34s (B+ around 450V or so depending on wall voltage and bias) it should be 220k in series with the diode in the bias circuit and 56k in line with the bias pot, plus 220k for the grid load/bias return resistors. For 6550s 150k, 47k, 150ks respectively. The R in series with the bias diode gives more negative voltage if you go lower, the R in series with the bias pot gives less as you go lower, and the grid load/bias return Rs won't effect the bias voltage but can have influence on the stabilty of the outputs. The R in series with the bias pot and the bias pot itself make up a resistance from C- (the bias voltage) to ground--the less the resistance is the less negative voltage it gives. The fixed R (47k or 56k) in series with the bias pot makes sure the adjustment range is restricted and you don't short out the C- to ground running the tubes too hot and damage them. The R in series with the diode looks kind of odd maybe brown red (sort of looks orange but maybe it's actually yellow?). If it was 120k, I think could mean that it has a lower B+ PT as found in JMP 50Ws so that was the case 120k would make sense (bit lower B+, bit lower negative bias voltage range). If you want to know what the resistances are, the best thing is to just measure them (amp unplugged, voltage drained, etc.).
-
- Senior Member
- Posts: 386
- Joined: Fri May 12, 2006 11:34 pm
- Location: Oahu HI
- wdelaney72
- Senior Member
- Posts: 1619
- Joined: Sun Jan 22, 2006 9:04 pm
- Location: Chicago Suburbs
-
- Senior Member
- Posts: 386
- Joined: Fri May 12, 2006 11:34 pm
- Location: Oahu HI
- Mr Crumb
- Senior Member
- Posts: 357
- Joined: Thu May 25, 2006 5:31 pm
- Location: East Coast
-
- Senior Member
- Posts: 386
- Joined: Fri May 12, 2006 11:34 pm
- Location: Oahu HI
- wdelaney72
- Senior Member
- Posts: 1619
- Joined: Sun Jan 22, 2006 9:04 pm
- Location: Chicago Suburbs
By design EL34 and 6550 differ in:
1) Plate Dissipation
2) Screen Grid Voltage
3) Contro Grid Voltage
4) Cathode Current
6550 are a different, more robust tube. They run differently. When Marshall opted to switch USA models to 6550, they adjusted the bias circuit accordingly. Not because they wanted to as it was an added expense, but because it's required for the tube to be biased properly. KT88 are the same as 6550, so a jump from EL34 to KT88 would need the same changes.
1) Plate Dissipation
2) Screen Grid Voltage
3) Contro Grid Voltage
4) Cathode Current
6550 are a different, more robust tube. They run differently. When Marshall opted to switch USA models to 6550, they adjusted the bias circuit accordingly. Not because they wanted to as it was an added expense, but because it's required for the tube to be biased properly. KT88 are the same as 6550, so a jump from EL34 to KT88 would need the same changes.
-
- Senior Member
- Posts: 386
- Joined: Fri May 12, 2006 11:34 pm
- Location: Oahu HI
- wdelaney72
- Senior Member
- Posts: 1619
- Joined: Sun Jan 22, 2006 9:04 pm
- Location: Chicago Suburbs
I'll have to defer to Larry, Dan, Steve or some of the other true electroncs experts. I believe it was Dan (Billy Batz) who I think sent me the info I listed earlier on the changes needed.Marshall Growl wrote:All I did was change from a 56k to around 70k. But, on my other thread, Wdelaney says you should do more than this, so it has me a little worried. They'r in range now to run them at about 28 watts at idle. So, I've gotte some mixed oppinions on this.
- novosibir
- Senior Member
- Posts: 4654
- Joined: Tue Nov 22, 2005 2:32 pm
- Just the numbers in order: 7
- Location: Nuernberg, Germany
- Contact:
You still have to swap the 220K bias feed R's with 150K's
Larry
Larry
The fault almost always is sitting in front of the amp 
Larry's Website now with included Pix's Gallery

Larry's Website now with included Pix's Gallery
-
- Senior Member
- Posts: 386
- Joined: Fri May 12, 2006 11:34 pm
- Location: Oahu HI
- MajorWatt
- Senior Member
- Posts: 345
- Joined: Sun Oct 15, 2006 11:45 pm
- Location: S.F. Bay Area
On my '68 Metro board, I just changed the bias voltage input resistor to a 4.7K and I was able to get between -70 and -80 vdc.
Reverb Nation/Black Sunday Dream
SoundClick/Black Sunday Dream
http://www.blacksundaydream.com/" onclick="window.open(this.href);return false;
SoundClick/Black Sunday Dream
http://www.blacksundaydream.com/" onclick="window.open(this.href);return false;
- novosibir
- Senior Member
- Posts: 4654
- Joined: Tue Nov 22, 2005 2:32 pm
- Just the numbers in order: 7
- Location: Nuernberg, Germany
- Contact:
Just to make it clear once again, based on Marshall's technical informations for use of 6550A/KT88 tubes in the 1959 and 2203 you have to change alltotal 4 resistors:
- both the bias feed resistors from 220K to 150K
- in the bias supply the 27K from the white cable to the diode to 15K
- the 56K in front of the trim pot to 47K
That's all and you will have the needed bias adjustment range and the safety margin for the output tubes, to avoid blocking distortion.
Larry
- both the bias feed resistors from 220K to 150K
- in the bias supply the 27K from the white cable to the diode to 15K
- the 56K in front of the trim pot to 47K
That's all and you will have the needed bias adjustment range and the safety margin for the output tubes, to avoid blocking distortion.
Larry
The fault almost always is sitting in front of the amp 
Larry's Website now with included Pix's Gallery

Larry's Website now with included Pix's Gallery
-
- Senior Member
- Posts: 386
- Joined: Fri May 12, 2006 11:34 pm
- Location: Oahu HI