WAD 300B PP Rebuild Advice Needed.

[ZakB]

Hi Richard, John and Greg,

Thank you all for your helpful suggestions and links to a lot of history!

Based on everything you've suggested, the decision is made, I will completely strip down the amp and will return the configuration to standard with the addition of:

1) The input section built on tag boards so that I can optimise the earthing arrangement.

2) Lift the earthing correctly.

3) Wire the PP 300b heaters in-phase.

4) Lower the 300b plate dissipation if needed.

I'd like to see how it is without input DC heaters first as Richard, if I am reading the links you sent correctly, you managed to achieve low levels of hum with AC heaters?

Greg, thanks for the pictures of a couple of nicely built amps :-). Am I right in thinking that neither of these had a mains switch on the front panel? I ask, as I've never understood this aspect of WAD kits - taking hum inducing mains from the rear, to the switch at the front and then back again to the transformer. The 300b layout also makes it difficult to keep the wires in the chassis corners as the left channel interstage transformer is in the way. I think i'll machine up a blanking plate and do away with the front switch....

I'll detail the rebuild here but it may take a while John. With two young children and a busy job, what used to take a weekend now seems to take millennia!

As an aside, Richard and Greg, I hope you don't mind me asking, but why did you move your amps on?

[Richard]

Hi Zak, that looks a good plan.

Mine came with DC heaters for the small valve(s?) which I kept so I don't know about the original AC heaters.

I found a couple of pics of the original and changed earthing routes on mine - at the time I borrowed Greg's pic to draw on but you'll see the earthiest point is the large pink splodge so that's where I took the leads to avoid modulation.

I've also found a pic of mine in final guise attached,

it has Soniqs coupling caps http://www.world-designs.co.uk/capacitors.htm
and you'll see the purple Oscon bypass caps piggy backed onto the 1K first stage cathode R3. I didn't use fb, and preferred cathode bias to battery bias, so removed R4 as well and put R3 right across the holes which gave a nice long couple of legs to solder the Oscons to. I listened on/off and kept them.

Yours has battery bias I see so you might want to remove or change the batts for the rebuild as they may be tired now.

[Greg.]

Originally Posted by ZakB

Greg, thanks for the pictures of a couple of nicely built amps :-). Am I right in thinking that neither of these had a mains switch on the front panel? I ask, as I've never understood this aspect of WAD kits - taking hum inducing mains from the rear, to the switch at the front and then back again to the transformer. The 300b layout also makes it difficult to keep the wires in the chassis corners as the left channel interstage transformer is in the way. I think i'll machine up a blanking plate and do away with the front switch....

As an aside, Richard and Greg, I hope you don't mind me asking, but why did you move your amps on?

Yes, when a group of us built clones, we changed so the power switch was at the back in line with other WAD amps at the time. I got seven chassis made in a single order and this made them individually affordable. The last chassis went to Tony Moore. By then we had run out of pattern PCB’s, he se Tony’s tag strip build. Incidentally, one earthing mod we introduced to the main PCB was to cut the earth track and then run two separate wires to star earth. I can’t remember now the location of that cut, but Richard may be able to help. We also had 6mm brushed and anodised faceplates made to bring the appearance in line with WAD amps at the time.

I sold mine mainly due to domestic pressure because I had too many boxes in the living room and valves are not grandchild friendly. I moved to a fully integrated Devialet which has streaming board, phonostage, dac, pre and power all in one case. Also multiple programmable options to suit any need I have and all in one slim and attractive chassis. Sound is fantastic! I do still use an external hybrid phonostage with a valve output because I prefer the performance over the Devialet option. Other considerations was the repeated need to buy new output valves every couple of years and occasional GZ37 which are silly money now if Mullard is the preference.

[Richard]

Hi Greg, and All,
I didn't cut the pcb track, just disconnected the pcb ground from the 300B ground, then wired the neg legs of the C5 caps to the ground splodge as you may be able to see by the pink line and black wires in the photos.
I suppose my arrangement was a mixture of bus bar and star and you could break tracks and take 4 separate grounds back from the pcb from each small valve as well (edit, just seen I did the speakers separately, the idea was to stop large ground currents modulating small)

[ZakB]

Hi Greg and Richard,

Thanks again for all your feedback ;-)

I'll follow your star earthing arrangement Richard. The amp does have battery bias, although this was one of the reasons it wasn't working when I got it. 2 of the cells were completely shot and wouldn't accept any charge.

Greg, I completely understand the grandchildren+valves issues, although in my case it is my own children. One will do exactly as you ask her and not touch them, the other will do precisely the opposite :-).

Will keep you all posted on (likely slow) progress!

Zak

[Richard]

Good luck and let us know how it goes Zak. Valve amps are inherently untidy and personally I'd just go for a tidy-up

[ZakB]

Hi All,

The rebuild is pretty much complete. So far it's had:

- A complete strip down and individual test of all components that were to be re-used.

- The chassis and hardware have been reconstructed with A2 stainless cap head bolts.

- The input board was shot and as John C pointed out, could have it's earthing improved. I designed a new one with separate channel earths and the revised earth sequencing to minimise larger currents modulating smaller ones. I went for traced earths in 2oz copper rather than a ground plane.

- The board was designed with resistive bias and no feedback compnents. I did experiment with feedback but as others have reported found it produced a poorer sound.

- The whole earthing arrangement was revised as Richard suggested. The input cabling is sheathed twin core with the sheath earthed at the amp input only.

- The output valve dissipation has been lowered.

The result is an amp that sounds fabulous and pairs very well with my sensitive (95dB@1m@2.83V) Kelly KT3's. It's everything I'd hoped for and I'd like to thank all of you who have helped me with advice on the rebuild. :-)

As Matthew warned me it would, it does still hum a little. Measured with a multimeter, there is 9mV on both channels. It is not intrusive, but I'm working to eliminate it as much as possible and hit that 2mV target Richard and Greg achieved. Early investigations and signal tracing with a scope suggest it is not a ground loop but coupled from somewhere. I'll report more after further high voltage probing.....

Thanks again and hope you are all well,

Zak

Amp Overview Alternate by Zak Barrett, on Flickr

[IMG]Amp Inside by Zak Barrett, on Flickr[/IMG]

[IMG]Input Board Component Side by Zak Barrett, on Flickr[/IMG]

[IMG]Input board Valve Socket Side Alternate by Zak Barrett, on Flickr[/IMG]

[Phil Y]

That is a great rebuilt Zak, a huge improvement on how it was.

Phil.

[Greg.]

+1

[John Caswell]

Hi Zak,
Very nice job.
Couple of things that may help.
On r/h side of amp in picture the yellow and purple leads of the driver tx grid feed and the 300B anode feed are too close together, that can cause oscillation which could manifest itself as hum.
As the 300b grids are fed via the tx it is easy to temporarily fit a two fat (1.0µF) capacitors between the grids and cathode bias earths to see if hum is eliminated. This may help to pinpoint where the hum comes in.
As GZ37 are now very expensive you could try a 6D22S in its place as all it is doing is acting as an HT delay. It would be easy to knock up an Octal to B9D base adaptor.The 6D22S runs very well on 5V even though its heater is rated at 6.3V.
I use them in my GEC 88/50 amps without any problem.

John