Generally, if you have a 2-chassis construction, I would suggest that the chokes are mounted with the transformer, if possible. The chokes are unshielded, and can spill EM fields on to the signal parts.
The best way to check the 'umbilical' for risk of noise:
- model all of the power supply in PSUD2. The look at the peak-current into the capacitor on the signal-chassis. 100/120Hz capacitor-recharge current will cause coupling to other cables in the bunch.
1mA peak is probably OK, 20mA peak is probably enough to get some coupling, at a rough guess. You can try adjusting the balance of capacitance on the power chassis vs. the signal chassis, to get improvement.
If there are recharge-currents in the cable, twisting the +/- supply-pair is more effective than shielding, and twist the 'victim' cable-pair, too.
Shielding the cable pairs is needed, if there is voltage-noise (Ripple-voltage).
Signal Current:
Only A small capacitor on the signal chassis might mean that there is some signal-current in the B+ cable. A shunt-regulator is a good fix for this, since it makes the supply-current completely static dc.
But with low-current valves like the 26, this is unlikely to be a problem.
Sorry to interrupt - I've got pair of UTC HA133 for sale. Please PM for details or check "swap meet" section.
Peak? On HT I have min;max;diff;mean;RMS. Diff is 430.66n and RMS is 157.34n. So less than 1mA any way I read it?
1. Plot the value of (say) I(C3) if C3 is the cap on the signal chassis.
2. The current into this cap is the current in the cable.
3. Look at the plot of current I(C3), and check that it is stable - i.e. not still ramping up from power-ON. If not, increase the Reporting Delay.
4. Read the peak current as DIFF. (or just look at the plot to get a rough idea)
2. The current into this cap is the current in the cable.
3. Look at the plot of current I(C3), and check that it is stable - i.e. not still ramping up from power-ON. If not, increase the Reporting Delay.
4. Read the peak current as DIFF. (or just look at the plot to get a rough idea)
Gotcha. C3 is on the signal chassis. And L2, just before C3 is also on the signal chassis. Diff on that is 6.7n.
Last edited:
I've put the AZ1 meshplate back in, and somehow it reduces the annoying sibilance that the AZ31 seemed to emphasise. Oh well, I guess it stays then. 🙂
BTW, I should mention that this is with the 01A/gyrator, but I did experience very similar differences with the 26.
Hi Guys
Little advice pls, is it imperative that the Psu must
be seperated to avoid Hum ?
Thks
Little advice pls, is it imperative that the Psu must
be seperated to avoid Hum ?
Thks
I wanted to build my preamp in one chassis only, but disregarding hum, the size and weight of it all was just too much. You need room to change parts and tweak the sound. I would say 2 chassis minimum or 3 smaller ones for signal, HT supply and Filament supply. You will not regret it if it is your first build. Trust me.
Thanks for the reply. I have the parts layed out
on a plywood board & it's not big & heavy. My only
concern is how to lay out the transformer & chokes
to prevent hum. What parts are you using that's making
it big & heavy ? Im going to use 3 chokes , first from
from rectifier then split Dc to 1 choke for each
channel. Will see how it turns out on the ply board
Thks again
on a plywood board & it's not big & heavy. My only
concern is how to lay out the transformer & chokes
to prevent hum. What parts are you using that's making
it big & heavy ? Im going to use 3 chokes , first from
from rectifier then split Dc to 1 choke for each
channel. Will see how it turns out on the ply board
Thks again
I use a mix of Weiss, Lundahl and Hammond. 3 power transformers and 5 chokes. Not that much but it takes up som real estate. At least mine does. The most important aspect in avoiding hum is the conctruction of the filament supplies, and that you have short leads from the rectifier to the first capacitor. I use choke input myself to avoid most of the hefty current pulsing.
Please load up some pictures when you are finished breadboarding so we can have a look at it.
Please load up some pictures when you are finished breadboarding so we can have a look at it.
Thks for the reply. Yes will do so when Im done.
Right now what's missing are the rectifiers
just bought a couple of Telefunken Az1's off
Ebay. Was thinking of other rectifiers but
after asking as well as some researching
might as well bite the dust & be done with
it.
Thanks again
Right now what's missing are the rectifiers
just bought a couple of Telefunken Az1's off
Ebay. Was thinking of other rectifiers but
after asking as well as some researching
might as well bite the dust & be done with
it.
Thanks again
Btw I presume that your using Rod's filament regs as well ?
In reality I have a different usage for this line stage. If it does
not work out I will relegate it back to a line stage
Thanks again
In reality I have a different usage for this line stage. If it does
not work out I will relegate it back to a line stage
Thanks again
You can trust me too! I now use 3 chassis for everything - signal, HT and filaments. Each to switch things around. We're all experimenters here.
I use two, and that means the PSU chassis for the 26 is now the PSU for the 01A, with just one resistor change. At least for the moment. I intend to add a switch on Monday so I can change preamps in a few seconds.