Good VFM? The PANDA headphone amp. I'll say so!

[Deleted]

The 100Hz AC ripple will be filtered best with a higher zener current as the current through the zener increases so will the dynamic resistance drops (and so will the AC ripple)
Original the zener current will be around 6mA, with the FET around 2mA.
BUT since it acts as a current source the ripple won't be entering the zener (as the current is constant) and this will at least compensate.
A 0.4W instead of 1.3W zener might help too in the case of the FET.

[Deleted]

Nov 8, 2010 6:11:34 GMT @solderdude said:

The 100Hz AC ripple will be filtered best with a higher zener current as the current through the zener increases so will the dynamic resistance drops (and so will the AC ripple)
Original the zener current will be around 6mA, with the FET around 2mA.
BUT since it acts as a current source the ripple won't be entering the zener (as the current is constant) and this will at least compensate.
A 0.4W instead of 1.3W zener might help too in the case of the FET.

Frans
Are you saying that the Zener Diode they are using is a 1.3W device with only 6mA through it ?
Even the 18V 1W 1N4746A requires 14mA (typically) through it to have 18V.
Alex

[leo]

I normally use J309's and stick a resistor on source adjusting it to required current but thought it maybe worth a try with the leftover 246's rather than waste them. (source straight to gate)
Output voltage is the same as with the 1k5's, just a little under 18v. I got just over 5mA across the 1k5, I have some J309's so will try those.

[Deleted]

Nov 8, 2010 9:11:00 GMT leo said:

I normally use J309's and stick a resistor on source adjusting it to required current but thought it maybe worth a try with the leftover 246's rather than waste them. (source straight to gate)
Output voltage is the same as with the 1k5's, just a little under 18v. I got just over 5mA across the 1k5, I have some J309's so will try those.

BTW, Mr. Marmite, I just had a nice snack of Vegemite on toast.
Everything goes better with Vegemite !

[Deleted]

I decided to make some 3-4mA CRD's and try replacing the zeners bias resistors.

hi Leo

time to show my ignorance but what's a CRD?

I'm new to the Solid State thing so i have some learning to do

take care

[PinkFloyd]

Panda is lashed up and running in , trimmers set for 1v5

Jesus man, you don't hang around do you? It only arrived on Saturday!

[Deleted]

Nov 8, 2010 7:35:22 GMT @sandyk said:

Nov 8, 2010 6:11:34 GMT @solderdude said:

The 100Hz AC ripple will be filtered best with a higher zener current as the current through the zener increases so will the dynamic resistance drops (and so will the AC ripple)
Original the zener current will be around 6mA, with the FET around 2mA.
BUT since it acts as a current source the ripple won't be entering the zener (as the current is constant) and this will at least compensate.
A 0.4W instead of 1.3W zener might help too in the case of the FET.

Frans
Are you saying that the Zener Diode they are using is a 1.3W device with only 6mA through it ?
Even the 18V 1W 1N4746A requires 14mA (typically) through it to have 18V.
Alex

Yes they are using 1N4746A 1.3W types with 6mA running through it.
They have about 17.5 Volts on them because of this.

I would use 0.4W types myself for this reason.

They can be used from a few mA's though not optimal (higher dynamic resistance) and 50% below the typical value is just acceptable.

[Deleted]

hi Leo

just had a quick look with google and found this;

Current Regulating Diodes ''provide isolation between the high and low-current sections of the power rails. This means that any ripple put on the rails by the high-current output section will be greatly attenuated''

i think i understand

take care

[Deleted]

CRD is a single device with 2 legs and is in fact a CCS constant Current Source that you can buy in certain fixed values.
a j-FET can also be used as a CCS by connecting the gate to the source but you won't know what current you will get exactly until you measured it.
It will be within a certain range.
You can also make CCS with transistors and a reference diode or other reference voltage such as a LED (as used in the Bravo thingies)

All these parts provide a constant current and will keep this same current no matter what voltage is applied.
At least within a range, their minimum voltage and maximum one.

[leo]

Nov 8, 2010 11:02:16 GMT PinkFloyd said:

Panda is lashed up and running in , trimmers set for 1v5

Jesus man, you don't hang around do you? It only arrived on Saturday!

I appreciate you sending it so fast (thanks again) , it seemed only right to get it lashed up during the weekend I was out at York Saturday daytime so knocked it up Saturday night

I didn't want to comment too much on SQ just yet because I feel it needs plenty more run in time but so far I must say its sweet yet very detailed , compared to say the SCHA the Panda is a thicker/weightier sounding amp, the bass is insane no hint of any HF nasties even with the HD250.

If I can dig the CKIII out I'll see how it compares to that after its had some more run in

[leo]

Nov 8, 2010 11:40:49 GMT @solderdude said:

Nov 8, 2010 7:35:22 GMT @sandyk said:

Frans
Are you saying that the Zener Diode they are using is a 1.3W device with only 6mA through it ?
Even the 18V 1W 1N4746A requires 14mA (typically) through it to have 18V.
Alex

Yes they are using 1N4746A 1.3W types with 6mA running through it.
They have about 17.5 Volts on them because of this.

I would use 0.4W types myself for this reason.

They can be used from a few mA's though not optimal (higher dynamic resistance) and 50% below the typical value is just acceptable.

Hopefully the guys behind the kits is reading this and may start supplying the lower wattage type zeners . I intended to get some from Maplin at the weekend, unfortunately they are out of stock.

My J309's range between 12mA and 14mA so I may make some CRD's with those, its no problem to add a resistor for adjusting the current so they all match, the 246's I used was 3.6 to 3.8mA's

[leo]

Nov 8, 2010 13:00:27 GMT @shaun said:

hi Leo

just had a quick look with google and found this;

Current Regulating Diodes ''provide isolation between the high and low-current sections of the power rails. This means that any ripple put on the rails by the high-current output section will be greatly attenuated''

i think i understand

take care

Good man
I've used CRD's in similar circuits with positive results

[PinkFloyd]

I've fired in a couple of 33R output resistors and I think these are a better match for the HD-250... as you say, Leo, the bass is insane! It's becomming more and more obvious as the weeks pass that this is amp will respond well to many tweaks. Tomorrow I'm going to try 9.1R Kiwame output resistors.

[Deleted]

hi Frans and Leo

thanks for the info it's helped ;D I've only used ccs (Depletion Mode Mosfets)on the anodes of tubes and like the clarity the can bring.

it could be interesting to try the CRD thing with .4W zener to see what it does to the sound if anything as i say I'm new to all of this but it's interesting to do something different, IMHO

take care

[leo]

Nov 8, 2010 19:10:06 GMT PinkFloyd said:

I've fired in a couple of 33R output resistors and I think these are a better match for the HD-250... as you say, Leo, the bass is insane! It's becomming more and more obvious as the weeks pass that this is amp will respond well to many tweaks. Tomorrow I'm going to try 9.1R Kiwame output resistors.

Thanks for the update Mike, pretty easy values to source too, I've got some normal 0.7w ones I can use for trying out, which ever goes best with the HD250's

[leo]

Nov 8, 2010 23:20:17 GMT @shaun said:

hi Frans and Leo

thanks for the info it's helped ;D I've only used ccs (Depletion Mode Mosfets)on the anodes of tubes and like the clarity the can bring.

it could be interesting to try the CRD thing with .4W zener to see what it does to the sound if anything as i say I'm new to all of this but it's interesting to do something different, IMHO

take care

Hi Shaun,

Some time ago I knocked up a valve output stage for a dac I was fiddling with, I tried Gary Pimms CCS on the anodes and remember the improvement it made, certainly worth the effort.

I've now set 4 x J309's at 10mA's and fitted those in the Panda's regulation instead of the 246's, the stock zeners now measure 18v well actually 3 are 18v and the other is 18v1

Cheers,
Leo

[Deleted]

Hi Leo

AHHH yes Mr pimms is the main man for tube ccs, I've used some of his before going back to plate load chokes where possible.

I've ordered some j309 from farnell so waiting on them to turn up.

oh yes i used a 9v battery set the current on my Dep mosfets, is it the same again with the j309?

any chance of suggesting a useful range of resistors to have to hand to set the j309 for 10ma?

questions questions ;D

much appreciated

take care

[PinkFloyd]

Nov 9, 2010 12:11:11 GMT @shaun said:

AHHH yes Mr pimms is the main man for tube ccs

[Deleted]

Nov 9, 2010 12:26:07 GMT PinkFloyd said:

Nov 9, 2010 12:11:11 GMT @shaun said:

AHHH yes Mr pimms is the main man for tube ccs

Only if it's a pink one

[Deleted]

Nov 8, 2010 19:10:06 GMT PinkFloyd said:

I've fired in a couple of 33R output resistors and I think these are a better match for the HD-250... as you say, Leo, the bass is insane! It's becomming more and more obvious as the weeks pass that this is amp will respond well to many tweaks. Tomorrow I'm going to try 9.1R Kiwame output resistors.

Mike,

Do these resistors just improve the Bass with the 250`s, or are there benefits throughout the sound spectrum.??

Ian`s (toad) Panda is now finished, i will keep it on test this week and send it to him Monday, sounds great with 100 ohm resistors +701`s, another happy bunny with a Panda, i hope.

[toad]

Nov 9, 2010 13:02:35 GMT @heliharris said:

Ian`s (toad) Panda is now finished, i will keep it on test this week and send it to him Monday, sounds great with 100 ohm resistors +701`s, another happy bunny with a Panda, i hope.

I'm sure I'll be doing Bugs impressions within minutes of getting my hands on the Panda Mick. Thanks very much for building it for me mate.

Ian

[leo]

Nov 9, 2010 12:11:11 GMT @shaun said:

Hi Leo

AHHH yes Mr pimms is the main man for tube ccs, I've used some of his before going back to plate load chokes where possible.

I've ordered some j309 from farnell so waiting on them to turn up.

oh yes i used a 9v battery set the current on my Dep mosfets, is it the same again with the j309?

any chance of suggesting a useful range of resistors to have to hand to set the j309 for 10ma?

questions questions ;D

much appreciated

take care

Hi Shaun,

A 9v battery is fine for measuring the J309's, on my vero made jig I added a 1k trimmer on the source for adjusting the current, the trimmer is then measured after the current is set and I use nearest value fixed resistor (seen below)

If your bored and wanting to try the CRD's be sure they are connected the right way around when soldered into the Panda or you'll have problems

I've now tred the J309's set for 7mA's so its closer to what was had with the stock fitted resistors

For 7-8mA resistors was 68R to 75R (with my batch) and 33R to 47R for 10mA, obviously there will be a variation between batches



Cheers,
Leo

[PinkFloyd]

What's she sounding like Leo? Do you think you'll bother with an enclosure?

[leo]

Nov 9, 2010 21:00:09 GMT PinkFloyd said:

What's she sounding like Leo? Do you think you'll bother with an enclosure?

I'm liking it , its weird though because its giving the hint it has a lot more to offer, something somewhere is just slightly holding it back if that makes sense?

Will I bother with an enclosure? for sure yes

[leo]

It really benefits from those few days run in I'm still not going to change any caps yet and will give another couple of days running in, going from past experiences Nichicons can change quite dramatic after fully burnt in, being half run in may be whats giving the impression of being slightly held back