6k6 3 watt build

reflux

TDPRI Member
Joined
Jan 7, 2016
Posts
56
Location
MT
So I put together this little amp from Terry at his youtube channel DLAB.
Doing a voltage check without any tubes
I have 316vac on each leg of the xfmr,

141vdc on each of the diode legs before they're connected to the first node of the filter cap.
Seems ok up to here, once both diodes connect to the first filter cap I have 443vdc,
same for the second node, and 441vdc on the 20uf third node.
I know voltages should be higher without a tube load but these seem very high to me.
I installed the 6k6 and brought the voltage up slowly with a variac until I reached about 280 volts, this was at an input voltage of only 70vac. I shut it down there.

Ideas? Appreciate any and all help you can throw my way. Thanks much.



This pic may be hard to read so here's the youtube address the schematic is at 3:37:



IMG_0961.JPG








REPLY
 

Jon Snell

Tele-Afflicted
Joined
Aug 31, 2015
Posts
1,299
Location
Jurassic Coast, Dorset. Great Britian.
Your mains transformer is ?
Output from an AC source through a full wave rectifier should be 1.414 times the AC feeding it.
You mean you have 443v on the diode cathodes?
Do you mean a 6K6G?
It is unlikely a 6K6, that would have been a metal cased version and I don't think any were produced.
 

reflux

TDPRI Member
Joined
Jan 7, 2016
Posts
56
Location
MT
Thanks for the reply Jon.
The xfmr is unknown and used.
The voltage specs are: 315-0-315, 6.3vac, 5vac
Yes there is 443v on the diode cathodes and it is a 6k6g.
So the RMS factor puts the voltage at an appropriate level it seems.
So would this be normal for the plate voltages(without tubes)to be carrying this higher voltage?
If so, perhaps increasing the dropping resistor values would help?
Thanks again
 

andrewRneumann

Tele-Afflicted
Joined
Mar 22, 2020
Posts
1,802
Location
Cincinnati, OH, USA
Thanks for the reply Jon.
The xfmr is unknown and used.
The voltage specs are: 315-0-315, 6.3vac, 5vac
Yes there is 443v on the diode cathodes and it is a 6k6g.
So the RMS factor puts the voltage at an appropriate level it seems.
So would this be normal for the plate voltages(without tubes)to be carrying this higher voltage?
If so, perhaps increasing the dropping resistor values would help?
Thanks again

Why do I get the feeling that 240VDC is a typo or 330vac is a typo? They don’t really jive with each other do they?
 

NTC

Tele-Afflicted
Joined
Mar 17, 2003
Posts
1,204
What happens if you use ONE diode - a half wave rectifier. 315 * 0.707 = 222.7 VDC. Low, yes, but it should work to get started.

280/1.424 = 198 V. You need a transformer with about 200 to 210Vac on each leg w.r.t. the center tap.

6K6 - I like it when folks use different tubes in amps. Note that the Fender 6G15 reverb uses a 6K6 as the reverb driver. The drive circuit is ALMOST a proper single ended amp. The reverb pan is an 8 ohm load. The power supply is a half wave rectifier feeding a choke to clean it up.
 

reflux

TDPRI Member
Joined
Jan 7, 2016
Posts
56
Location
MT
Thanks for the replys.
So I went ahead and powered up slowly with the tubes in place as vampwizzard suggested. No fireworks but hot tubes. Node voltages: 382,373,349.
12ax plates @234ish, K-1.7
6k6 P-373, K-27
no redplating but that glass is real hot
Perhaps increasing the 10k dropping res to 22k would help some.
Yeah Andy those #s do seem a bit odd huh.
I may look and see if I have a 300-0-300 or less xfmr and if so maybe give it a shot.
 

Jon Snell

Tele-Afflicted
Joined
Aug 31, 2015
Posts
1,299
Location
Jurassic Coast, Dorset. Great Britian.
If you increase the 'dropping' resistor, are you sure what that means ... a resistor in series with a load reduces the current flow and if sufficient current is flowing, the resultant voltage also.
No current flow = no voltage reduction.
The schematic calls for 115 0 115 volts not 315 0 315.
I recomend you use the correct transformer. Half wave rectification is not really the correct way to go without adding a 10H choke and fine tuning the smoothing components and that would be more expensive than using a lower value transformer.
 

andrewRneumann

Tele-Afflicted
Joined
Mar 22, 2020
Posts
1,802
Location
Cincinnati, OH, USA
If you increase the 'dropping' resistor, are you sure what that means ... a resistor in series with a load reduces the current flow and if sufficient current is flowing, the resultant voltage also.
No current flow = no voltage reduction.
The schematic calls for 115 0 115 volts not 315 0 315.
I recomend you use the correct transformer. Half wave rectification is not really the correct way to go without adding a 10H choke and fine tuning the smoothing components and that would be more expensive than using a lower value transformer.

You meant 165-0-165. 👍
 

jumpbluesdude

TDPRI Member
Joined
Jul 15, 2012
Posts
43
Location
eau claire, wi
1. Check to see what a 6K6 can take as far as voltage in an RCA tube manual.
2. Add 20% and that is about where you can max out. Vintage tubes can take a lot, but better to not push it. I built an amp with a quad of 7408 (military grade 6V6) using a bandmaster PT putting out 470 and they can take it. Tim Caswell (yeah, that one) says they should be fine. RCA tube manual says 365 for 6V6 and Fender ran them at 415 in a DR. I didn't like them running that hot and reduced the voltage to 430.
3. You won't get red plating - that is usually a bias issue. The tubes will just short and likely take out your PT.
4. Spec a PT closer to your target voltage.
 

vampwizzard

TDPRI Member
Joined
Apr 28, 2019
Posts
46
Age
35
Location
Abingdon, VA
Thanks for the replys.
So I went ahead and powered up slowly with the tubes in place as vampwizzard suggested. No fireworks but hot tubes. Node voltages: 382,373,349.
12ax plates @234ish, K-1.7
6k6 P-373, K-27
no redplating but that glass is real hot
Perhaps increasing the 10k dropping res to 22k would help some.
Yeah Andy those #s do seem a bit odd huh.
I may look and see if I have a 300-0-300 or less xfmr and if so maybe give it a shot.
The original design used a 330 center tapped transformer, meaning each leg was 165. 165*1.414 is about 233 volts rectified (theory/ideal). As others have suggested, you’ll need a lower voltage power transformer.
 

reflux

TDPRI Member
Joined
Jan 7, 2016
Posts
56
Location
MT
Thanks everyone,especially Jon, who pointed out what should have been obvious to me, the 330vac. Wow, what a brain fart on my part. Alright then, on to checking xfmrs for a proper ac voltage. Still can't believe I read that wrong, especially since I've already built several larger amps from scratch. I think I'll spend more time on the guitar rather than amp building.

Thank you all again. I'll report back any progress or lack thereof when I can.
 

Lowerleftcoast

Poster Extraordinaire
Joined
Dec 30, 2019
Posts
5,501
Location
california
I can appreciate a Police Radio might have had low voltage.

Imo the 6K6g can take the higher voltages. What I see from your numbers the plate to cathode voltage is only 346V. About 10% over the 6K6G datasheet. I would definitely try the amp at this B+. I would decrease the voltage on the 12AX7. OMMV.

The other PT should give lower B+ so that is an option as well.
 

printer2

Poster Extraordinaire
Joined
May 24, 2010
Posts
8,405
Location
Canada
At the high voltage you are running at the tube wants a 10k primary, hope you do not have the 5k as in the schematic. Swap in a 6V6 and you should be fine. Mind you, if the tube does not throw sparks or melt it may take it. Calculate the dissipation.
 

reflux

TDPRI Member
Joined
Jan 7, 2016
Posts
56
Location
MT
Thanks for that headsup printer2.
So went ahead and changed out the first dropping resistor to 7.1k and the second resistor to a 18k. And because that plate voltage is so high on the 6k6 I did put in a 6v6. The dissapation with the 6k6 was over 10. Not too good for a tube with a max of 8.5 or so.
So no meltdowns or sparks.
Votages with the 6v6 are: Nodes; 391/310/276
6v6: P=392, K=21.7 820 Bias resistor @ 814ohms
So it works out to 10.45W Pd, at about 87% dissap

12ax7 plates are: pin1=180, pin6=187, K=1.4
Let it run for a half hour and all good so far. No extreme heat anywhere.
I'll run it thru some exercise and see what happens.
If anyone has more input I would appreciate it especially since I seem to be apt to overlooking the obvious at times.
Thanks again all.
 

Lowerleftcoast

Poster Extraordinaire
Joined
Dec 30, 2019
Posts
5,501
Location
california
I would think the typical 6K6G would use about a 1k (or so) cathode resistor to be in the ballpark. The suggested cathode resistor for a 6V6 is 470R. You can install a larger resistor to use the 6K6G and parallel it with another resistor when you use a 6V6. Put the parallel resistor on a switch if you want to switch back and forth between tube types.

FYI:
When calculating plate dissipation on a cathode biased amp, use plate to cathode voltage. You can measure directly form pin3 to pin8 or subtract the cathode voltage from the plate voltage.
The screen current is included in your measurement, so to be more accurate, take off about 5.5% or measure the screen current and subtract the actual amount. (The Robinette bias calculator takes off 5.5% when making the calculation.)

6K6G
6k6 P-373, K-27, 814R cathode resistance
Plate to cathode voltage 373-27=346

6V6
P=392, K=21.7 820 Bias resistor @ 814ohms
Plate to cathode voltage 392-21.7=370.3
 
Last edited:

reflux

TDPRI Member
Joined
Jan 7, 2016
Posts
56
Location
MT
Thanks for that insight LLCoast, an important calculation that I forgot about. I'll probably stick with the 6v6 but I may try out your parallel suggestion to see what happens time permitting. Have to tackle a cabinet/chassis for this now. The old portable radio cab and its narrow chassis is presenting a few too many size problems.
 

printer2

Poster Extraordinaire
Joined
May 24, 2010
Posts
8,405
Location
Canada
I would think the typical 6K6G would use about a 1k (or so) cathode resistor to be in the ballpark. The suggested cathode resistor for a 6V6 is 470R. You can install a larger resistor to use the 6K6G and parallel it with another resistor when you use a 6V6. Put the parallel resistor on a switch if you want to switch back and forth between tube types.

FYI:
When calculating plate dissipation on a cathode biased amp, use plate to cathode voltage. You can measure directly form pin3 to pin8 or subtract the cathode voltage from the plate voltage.
The screen current is included in your measurement, so to be more accurate, take off about 5.5% or measure the screen current and subtract the actual amount. (The Robinette bias calculator takes off 5.5% when making the calculation.)

6K6G
6k6 P-373, K-27, 814R cathode resistance
Plate to cathode voltage 373-27=346

6V6
P=392, K=21.7 820 Bias resistor @ 814ohms
Plate to cathode voltage 392-21.7=370.3
What about the 5k primary?
 

Lowerleftcoast

Poster Extraordinaire
Joined
Dec 30, 2019
Posts
5,501
Location
california
What about the 5k primary?
IDK if the OP is using a 5k primary/8 Ohm secondary as shown on the schematic. If he is, adding a second speaker for 16 Ohm on the 8 Ohm secondary should take care of the mismatch. Then again a tube amp has some leeway with an impedance mismatch, although that could have an audible effect. It is good to be cautious.
 




Top