PeeCeeBee V4 GB!

[shaan]

Hi Shaan,

Received my boards today, many thanks
Bruce

Hi Shaan, boards arrived today, nicely packed as well, thank you.
Neal

 

[shaan]

GB4 update:

NealJ - 2 PCBs + 4 MOSFETS - SHIPPED
helfred - 2 PCBs - SHIPPED
Assesears - 2 PCBs - SHIPPED
Amitkg - 2 PCBs + 4 MOSFETs - SHIPPED
thekingstallion - 2 PCBs + 4 MOSFETs - SHIPPED
Surfatwork - 2 PCBs - SHIPPED
hajj - 2 PCBs + 4 MOSFETs - SHIPPED
keesraats - 2 PCBs - SHIPPED
sasha70 - 2 PCBs + 4 MOSFETs - SHIPPED
phildas - 2 PCBs - SHIPPED
CitizenKane - 6 PCBs - (SHIP WITH V4H)
asuslover - 2 PCBs - (SHIP WITH V4H)
craydive - 6 PCBs + 12 MOSFETs - (SHIP WITH V4H)
John70267 - 2 PCBs - SHIPPING INVOICE SENT

 

[shaan]

GB4 update:

NealJ - 2 PCBs + 4 MOSFETS - SHIPPED
helfred - 2 PCBs - SHIPPED
Assesears - 2 PCBs - SHIPPED
Amitkg - 2 PCBs + 4 MOSFETs - SHIPPED
thekingstallion - 2 PCBs + 4 MOSFETs - SHIPPED
Surfatwork - 2 PCBs - SHIPPED
hajj - 2 PCBs + 4 MOSFETs - SHIPPED
keesraats - 2 PCBs - SHIPPED
sasha70 - 2 PCBs + 4 MOSFETs - SHIPPED
phildas - 2 PCBs - SHIPPED
asuslover - 2 PCBs - SHIPPED
John70267 - 2 PCBs - FULL PAID
CitizenKane - 6 PCBs - (SHIP WITH V4H)
craydive - 6 PCBs + 12 MOSFETs - (SHIP WITH V4H)

 

[shaan]

GB4 update:

NealJ - 2 PCBs + 4 MOSFETS - SHIPPED
helfred - 2 PCBs - SHIPPED
Assesears - 2 PCBs - SHIPPED
Amitkg - 2 PCBs + 4 MOSFETs - SHIPPED
thekingstallion - 2 PCBs + 4 MOSFETs - SHIPPED
Surfatwork - 2 PCBs - SHIPPED
hajj - 2 PCBs + 4 MOSFETs - SHIPPED
keesraats - 2 PCBs - SHIPPED
sasha70 - 2 PCBs + 4 MOSFETs - SHIPPED
phildas - 2 PCBs - SHIPPED
asuslover - 2 PCBs - SHIPPED
John70267 - 2 PCBs - SHIPPED
craydive - 6 PCBs + 12 MOSFETs - SHIPPED
CitizenKane - 6 PCBs - (SHIP WITH V4H)

 

[mexakin]

just a minor thing, cause i measured yesterday and was wondering about how much gain the PeeCeeBee would deliver from input to output signal.

It states about: AC gain = 22, which I could not explain myself from the schematic.

In the feedback path there is a voltage divider 2k2 and 100R, this means from general divider rules:
the feedback voltage is (2k2+100)/ 100 = 23
so I guess this is a minor slipup and it should be 22, or am i wrong, because of the 2 caps in the path, of course all of this is only AC coupled, and DC gain simply is unity = 1.

This gain difference of course does not affect any behaviour, it just for my understanding to clear it up, any maybe for anyone else of interest

I checked the newer V4H and there it is right, there are 4k7 in parallel in the feedback path which gives a AC gain of 24,5, which is also stated in the first page of the group buy.


Maybe a little about why i measured it and what i wanna achive:
I got my MAckie 1604 hooked to the Peecebee, and the Mackie gives out up to 30Vpp on its output audio signal path.

And I , dumb me , always wondered why i can only listen to hearable music when all the faders of the output where almost all the way down ...

since at unity gain, its something about 5V maybe, this of course translates to 5x23 [V] at the output, which either way sounds way too loud gets clipped and or heavily distorted.

1.question: How much signal can i apply max at the input, could it be that some restrictions from the transistors cut the input signal anyway beyond a certain level, maybe 4 V or I dont know 10V.

2.question: Or does the output of the peeceebee to the speakers limit the input signal, since it gets distorted and clipped after the gain levels up my input signal beyond what I apply as supply voltages?

3.question: can i play around with the gain myself (of course i can) but does it affect things like THD etc. heavily or does it maybe even better the THD in a positive way.
More feedback seems always more nice, and I would apply less feedback, or am I wrong, and the lower the gain the more feedback will get back to the input, is there a sweet spot of feedback signal vs THD performance.
Sadly I do not have a spectrum analyzer at hand to proof my ideas, or better said, to find out myself.

Cause i have to find out the kind of best signal to noise in my whole signal chain, from input signal over to the mackie mixer to the peeceebee, maybe its better to have the mixer set around unity gain and play the peeceebee with around 5Vpp input, but hten with my speakers at hand, i would ruin my ears it is all a bit loud.

Next thing on my mind, i wonder how Voltage translates to loudness levels overall.
Since right now i think( will have to proof tonight or this week) 10V at my 4Ohm speaker ( proofed with a sinusgenerator at the inputs of the mackie and then the whole signal chain under control) gives P=U²/R 25W electrical in the speaker which translates maybe 1% speaker efficiency to 25mW audio power.

and that is quite listenable if i recall right.

Anyway thanks for reading and mybehelping.
cheers,
seb

 

[CitizenKane]

Depending on the TechSpecs the Mackie 1604 has a Z of 120 Ω at the outputs and the PeeCeeBee a Z of 10kΩ at the inputs?

 

[shaan]

just a minor thing, cause i measured yesterday and was wondering about how much gain the PeeCeeBee would deliver from input to output signal.

It states about: AC gain = 22, which I could not explain myself from the schematic.

In the feedback path there is a voltage divider 2k2 and 100R, this means from general divider rules:
the feedback voltage is (2k2+100)/ 100 = 23
so I guess this is a minor slipup and it should be 22, or am i wrong, because of the 2 caps in the path, of course all of this is only AC coupled, and DC gain simply is unity = 1.

This gain difference of course does not affect any behaviour, it just for my understanding to clear it up, any maybe for anyone else of interest

I checked the newer V4H and there it is right, there are 4k7 in parallel in the feedback path which gives a AC gain of 24,5, which is also stated in the first page of the group buy.

Sorry for late reply. I was outside and saw your post and was going to reply but thought it would be better to get home then do it. There are some good questions asked.

In theory the gain should be 23 but in practice it's always less than that. Check the 1KHz scope shot I took a few minutes ago in my V4. Yellow trace is input and blue trace is output. 1V PP in and 22V PP out.

The same is true for V4H which shows 23.5. I think I mistakenly published the theoretical value for the bigger amp.

Maybe a little about why i measured it and what i wanna achive:
I got my MAckie 1604 hooked to the Peecebee, and the Mackie gives out up to 30Vpp on its output audio signal path.

And I , dumb me , always wondered why i can only listen to hearable music when all the faders of the output where almost all the way down ...

since at unity gain, its something about 5V maybe, this of course translates to 5x23 [V] at the output, which either way sounds way too loud gets clipped and or heavily distorted.

1.question: How much signal can i apply max at the input, could it be that some restrictions from the transistors cut the input signal anyway beyond a certain level, maybe 4 V or I dont know 10V.

It depends on two main factors, PSU voltage and load impedance.

With my 35V PSU, I noticed the amplifier clips at 28V with 8R load and at 25V with 4R load. So with 35V rails, peak input voltage for 8R load is 28/22=1.27V (around 900mV RMS), and peak input voltage for 4R loads is 25/22=1.13V (~800mV RMS).

With a higher PSU voltage the clip ceiling get higher so the amp can allow for a higher input voltage before clipping. Same happens when load impedance is higher.

2.question: Or does the output of the peeceebee to the speakers limit the input signal, since it gets distorted and clipped after the gain levels up my input signal beyond what I apply as supply voltages?

Any input that clips the output is an overload and should be avoided. I have clipped the output to within a couple volts of the rails, looking like almost a squarewave with a 10V PP sinewave input signal; but that's possible because there was no load connected. Any load at output and the clip ceiling comes down to a lower value.

Output level doesn't limit the input signal.

3.question: can i play around with the gain myself (of course i can) but does it affect things like THD etc. heavily or does it maybe even better the THD in a positive way.

If you want more gain then I suggest replacing the two 100R feedback resistors with 82R. You get close to 27 gain but expect a little sacrifices being made in the Class-AB THD amount. Bias it to 1Ampere Class-A and I think you'll never notice any problem with the sound.

More feedback seems always more nice, and I would apply less feedback, or am I wrong, and the lower the gain the more feedback will get back to the input, is there a sweet spot of feedback signal vs THD performance.

Yes. For every amplifier out there, a sweet spot exists. I have tried many feedback/gain configurations and 22 sounded like a very balanced sweet spot for 100mA Class-AB operation with awesome musical presentation. Of course one can lower the gain and increase feedback and attempt to make the amplifier even more linear (in measurements at least), but I seriously doubt it will sound any better.

Sadly I do not have a spectrum analyzer at hand to proof my ideas, or better said, to find out myself.

Cause i have to find out the kind of best signal to noise in my whole signal chain, from input signal over to the mackie mixer to the peeceebee, maybe its better to have the mixer set around unity gain and play the peeceebee with around 5Vpp input, but hten with my speakers at hand, i would ruin my ears it is all a bit loud.

An idea - a volume control followed by a unity gain buffer between the Mackie and PeeCeeBee. How about that?

Next thing on my mind, i wonder how Voltage translates to loudness levels overall.
Since right now i think( will have to proof tonight or this week) 10V at my 4Ohm speaker ( proofed with a sinusgenerator at the inputs of the mackie and then the whole signal chain under control) gives P=U²/R 25W electrical in the speaker which translates maybe 1% speaker efficiency to 25mW audio power.

and that is quite listenable if i recall right.

Anyway thanks for reading and mybehelping.
cheers,
seb

A 93dB sensitive driver needs half the power (i.e. 0.707 times the voltage) as a 90dB driver to sound equally loud (considering they both have the same coil resistance). Now, how much of input power in any driver is transferred to the air as acoustic waves and how much of it is dissipated in the voice coil as heat depend on a complex set of many factors like cone weight, cone area, surround and spider compliance, magnet strength, voice coil inductance etc. and the engineering of these parameters to create more efficient drivers is kind of a black-art much like RF electronics.

 

[mexakin]

thanks shaan for the quick reply, dont worry in future, i can wait up to the point that i forget that I even asked

some other open quesitons in my mind:

1. How do you measure distortion, since this is kind of a hard value to measure especially if it comes down to the real voltage noise, i assuem you are using your audio card input? just because your pictures seem rather from a PC screen than an good old HP spectrum analyzer

I am so looking forward to get my handy on something old, until then maybe i willl measure the peeceebee with athe little analog discovery, but honestl yi kind of lacking trust in that little device, some time ago i tried to measure a thing, maybe my signal generator with it, and it seems to switch some things internally, so the measurments are not consistent over various measurment settings, of course in the general it should be fine, but i am talking small noise stuff, anyway right now, i havent tried it out with trustable sources so just chitchat in here

and yes nice for stating the last thing about voltage and power translating to sound preasure. I got some cheapo speakers as my monitors i did buy used but they sounded quite alright in direct comparison to some others i had lying around, so no chance to get the SPL? is it calles like that, the sensitivity of the speaker, i will try out now and maybe write some findings later.

The idea with a pot wuold work but would involve some of my work first so i think i might try to attenuate the output signal directly form the macki that is working but maybe costs me some quality sound i will hear that soon i guess.

why i started the whole hting is i wanted to know which level of sound is putting out my Mackie so i can get the signal into an analog to digital converter to finally be able to record some music with it, so rather independent of the peeceebee which is only my monitor speaker driver.

thanks again.until next time

 

[vvs07]

I'd like to get four boards when available. Please add me to your list for next GB. Thank you!

 

[shaan]

thanks shaan for the quick reply, dont worry in future, i can wait up to the point that i forget that I even asked

some other open quesitons in my mind:

1. How do you measure distortion, since this is kind of a hard value to measure especially if it comes down to the real voltage noise, i assuem you are using your audio card input? just because your pictures seem rather from a PC screen than an good old HP spectrum analyzer

I am so looking forward to get my handy on something old, until then maybe i willl measure the peeceebee with athe little analog discovery, but honestl yi kind of lacking trust in that little device, some time ago i tried to measure a thing, maybe my signal generator with it, and it seems to switch some things internally, so the measurments are not consistent over various measurment settings, of course in the general it should be fine, but i am talking small noise stuff, anyway right now, i havent tried it out with trustable sources so just chitchat in here

and yes nice for stating the last thing about voltage and power translating to sound preasure. I got some cheapo speakers as my monitors i did buy used but they sounded quite alright in direct comparison to some others i had lying around, so no chance to get the SPL? is it calles like that, the sensitivity of the speaker, i will try out now and maybe write some findings later.

The idea with a pot wuold work but would involve some of my work first so i think i might try to attenuate the output signal directly form the macki that is working but maybe costs me some quality sound i will hear that soon i guess.

why i started the whole hting is i wanted to know which level of sound is putting out my Mackie so i can get the signal into an analog to digital converter to finally be able to record some music with it, so rather independent of the peeceebee which is only my monitor speaker driver.

thanks again.until next time

Hi Sebastian.

I use a 48:1 resistor network between the output of amplifier and ground. It comprises of a 47K and a 1K resistor connected in series. Output connects to 47K and ground connects to 1K. Signal for measurement is taken from joint of the two reaistors and fed to Lune-in of PC, sampled at 96KHz. Maximum measurable frequency is 48KHz, in reality it's 40KHz or so. For a 1KHz signal Ican measure 39 harmonics.

Thr cable that feeds to Line-in is a 6ft long unbalanced mucrophone cable with 2 layer dielectric and excellent shielding mesh. I noticed this cable contributes most to the low noise floor.

For higher SPL just go for a larger driver.

 

[shaan]

I'd like to get four boards when available. Please add me to your list for next GB. Thank you!

GB5 starts.

 

[hajj]

Boards and MOSFETS received.
Thank you Shaan!

 

[shaan]

 

[Kokanee]

Package from India received at my brother in laws place in Calgary. Thank you shaan for all your effort.

Myles

 

[shaan]

 

[surfatwork]

and she's singing....and how well!! Great job, Shaan. Thanks very much.

Test setup below, connected to a pair of cheap Boston Acoustics speakers. Sounds great even with these - great imaging and excellent vocals. Dead silent even with ear on speaker.



480mV & 470mV across VAS 47R resistors. 115mA on the MOSFETs. -14mV DC offset

Next is the cabinet - hmmmm!!

 

[shaan]

Hi Surfatwork.

Glad to know they're up and running and you like its sound. I tried my best to make the V4 Rev2 as good as it can be and your review assured me that I am in the right path.

Thanks a lot and happy listening.
shaan

 

[Shailesh]

Shailesh - 2 PCBs + components

 

[kpsthakur]

Nothing except BD139/140 replaced with MJE340G/350G and 2x1N4148 replaced with 2xMUR160.

Hi Shaan, my quest for BAV21 is still open...you are suggesting MUR160 for V4, can I use it in V4H???

 

[shaan]

Hi kpsthakur.

I have funally procured a bunch of the BAV21 diodes so if you need them just let me know.

The main two requirements from these two diodes in both amplifiers are very fast operation and very low reverse capacitance. So MUR won't be nearly as good as these tiny glass beads.