Ok cool. Thanks!
In the interest of understanding, am I correct in the following calcs and logic?
1. Iq for each fet = add Voltage across each Rs (Vrs) divided by the sum of Rs, (Vrs+Vrs) /.92 Ohms = ~130-140 mA ?
2. Assuming 1. above is correct then based on my measurements, the standard deviation for Iq = 3-4 mA , which is 2-3% . That seems good.
3. Iq for the entire channel = ~0.135a x 2 = ~ 0.27a per pair x 10 pairs = 2.7a ? Given stated 30w class A then the gate source voltage would be around 10v.
4. Output fets are common drain mode so current gain is how this makes power?
5. Seems like the header pins A B in mb are tied to the first set of source resistors near the ob ribbon cable header? I can test by measuring both at the same time.
Here are the calcs
In the interest of understanding, am I correct in the following calcs and logic?
1. Iq for each fet = add Voltage across each Rs (Vrs) divided by the sum of Rs, (Vrs+Vrs) /.92 Ohms = ~130-140 mA ?
2. Assuming 1. above is correct then based on my measurements, the standard deviation for Iq = 3-4 mA , which is 2-3% . That seems good.
3. Iq for the entire channel = ~0.135a x 2 = ~ 0.27a per pair x 10 pairs = 2.7a ? Given stated 30w class A then the gate source voltage would be around 10v.
4. Output fets are common drain mode so current gain is how this makes power?
5. Seems like the header pins A B in mb are tied to the first set of source resistors near the ob ribbon cable header? I can test by measuring both at the same time.
Here are the calcs
| Rseq | 0.94 | Ohms | ||||||||
| Vrs | Veq | Iq | Vrs | Veq | Iq | |||||
| back close to header | 67 | 131 | 139.36 | 67 | 141 | 150.00 | back close to header | |||
| 64 | 74 | |||||||||
| 68 | 124 | 131.91 | 70 | 145 | 154.26 | |||||
| 56 | 75 | |||||||||
| 65 | 124 | 131.91 | 65 | 138 | 146.81 | |||||
| 59 | 73 | |||||||||
| 68 | 125 | 132.98 | 66 | 138 | 146.81 | |||||
| 57 | 72 | |||||||||
| 65 | 123 | 130.85 | 64 | 133 | 141.49 | |||||
| front A quadrant L | 58 | 69 | front B quadrant R | |||||||
| Std | 3.21 | 3.41 | Std | 4.42 | 4.70 |
1.yes
2.yes
3. X150.5 is having how much N channel mosfets and how much P channel mosfets? each (N+P) pair is one current vertical path; so say that you have 10 vertical current paths per channel, sum Iq is 10*one vertical Iq, which you did measure on "bias" headers
"Given stated 30w class A then the gate source voltage would be around 10v." ...... clarify that
4. yes; OS is nothing more but also nothing less than current buffer
5.irrelevant; if you need to conduct setting, bias headers are convenient, as long you use connectors with wires for these headers........ in case of short made with probes nothing will happen except burnout of traces to headers
if you want to make comparison table of all source resistor (sag) voltages, there is only one way to do it
2.yes
3. X150.5 is having how much N channel mosfets and how much P channel mosfets? each (N+P) pair is one current vertical path; so say that you have 10 vertical current paths per channel, sum Iq is 10*one vertical Iq, which you did measure on "bias" headers
"Given stated 30w class A then the gate source voltage would be around 10v." ...... clarify that
4. yes; OS is nothing more but also nothing less than current buffer
5.irrelevant; if you need to conduct setting, bias headers are convenient, as long you use connectors with wires for these headers........ in case of short made with probes nothing will happen except burnout of traces to headers
if you want to make comparison table of all source resistor (sag) voltages, there is only one way to do it
I re-read papa's article on FET amps and think I understand now.
3. This concept I misunderstood. The gate source voltage sets the bias....
let me try the Class A Calc again: at idle Iq is current from + to - rail across n and p. with no ac voltage to gate. When gate begins to see the + swing in ac voltage all the current moves to N channel so 2x the 1.35a (for the 10 n+p) = 2.7a . Peak Class A power is I^2 *R = 2.7^2*8 = 58.52 w and Average Class A = 29.16 w (1/2 Peak)
5. yes I use connectors (GFA555 distortion led connectors fit) wired to terminal block so no possible way to short.
I will make a table for all source resistor sag. This time I will tilt the heatsinks off the chassis and wait for the temp to equilibrate.
Thanks!
3. This concept I misunderstood. The gate source voltage sets the bias....
let me try the Class A Calc again: at idle Iq is current from + to - rail across n and p. with no ac voltage to gate. When gate begins to see the + swing in ac voltage all the current moves to N channel so 2x the 1.35a (for the 10 n+p) = 2.7a . Peak Class A power is I^2 *R = 2.7^2*8 = 58.52 w and Average Class A = 29.16 w (1/2 Peak)
5. yes I use connectors (GFA555 distortion led connectors fit) wired to terminal block so no possible way to short.
I will make a table for all source resistor sag. This time I will tilt the heatsinks off the chassis and wait for the temp to equilibrate.
Thanks!
I'll be back
Domani
Flue is still having my brain now, so can't check your calculus of A Class envelope
just remember - bridge amp is practically two normal/"SE" amps in series, where load (speaker) is nested between them
voltage envelope is what is doubled comparing to just one "SE" amp, while current envelope is same as for one "SE" amp
meaning - for class A envelope, you count on Iq of just one halve (phase) of amp, and double it to get sum A Class current envelope, exactly the same as amp is not even bridged
everything above that, proper amp, you can weld, but you can't call it "welded with A Class welder"
conclusion, I can write BS even with fuzzy brain, but I suck with math
Domani
Flue is still having my brain now, so can't check your calculus of A Class envelope
just remember - bridge amp is practically two normal/"SE" amps in series, where load (speaker) is nested between them
voltage envelope is what is doubled comparing to just one "SE" amp, while current envelope is same as for one "SE" amp
meaning - for class A envelope, you count on Iq of just one halve (phase) of amp, and double it to get sum A Class current envelope, exactly the same as amp is not even bridged
everything above that, proper amp, you can weld, but you can't call it "welded with A Class welder"
conclusion, I can write BS even with fuzzy brain, but I suck with math
Attachments
sebtdi, True but I believe in preventative maintenance. Testing will allow one to determine if that is needed. So far none, and that is what this analysis is about. That and educating myself about the circuit thanks to ZM. Peace.
ZM, no worries. Get well my friend. I have learned a lot with your guidance. Next week I may have some update.
So , Zm No need more test.. Sound is perfect. Went and tonight reached 54. Opened the door.. Turns out Jim is a neighbor so wrangled Iron pre boards out of him. It was tough and we dickerd on price for at least an hour after some hand drawn schematics and some mutual respect was achieved, well the price was Silly. Iron Pre SUSY was a thought you had and I listened. PS populated
Above hasty..To the point, at 25 ambient I'm 50 all day long idle. Pushing on A/B 54. Opened the door. Need to look at spec sheet for ops mofet but I do know Tc for ecx moves after 60. Peace. Hope things are well in your part of the world!
PS got a stockpile of Panasonic 100 ECX 3w mox 47 and another 100 82 for adcom ops.. can't forget your roots.. I've got lots of bjt carcasses yamaha P-2200 crown DC300a MKII from CU Boulder auctions. Likely those amps pulsed some part of the house sound for some pretty memorable performances. I digress.. looking forward to iron pre. Running PS audio 6.2 SE . It's just a burson V6 with a serviced noble np 25 and PSU . Burson too high gain so half input volume from perfect wave..DAC. And the passive SC-S2 came out to play tonight in front firing mode. An nht X1 of course dialed , PC -19 Dh440 monoblock doing duty on less than 70khz. Dunlavy. The man was one of the brilliants. Living the DAL legacy. That was PS ? Peace love and hi fidelity..
PS got a stockpile of Panasonic 100 ECX 3w mox 47 and another 100 82 for adcom ops.. can't forget your roots.. I've got lots of bjt carcasses yamaha P-2200 crown DC300a MKII from CU Boulder auctions. Likely those amps pulsed some part of the house sound for some pretty memorable performances. I digress.. looking forward to iron pre. Running PS audio 6.2 SE . It's just a burson V6 with a serviced noble np 25 and PSU . Burson too high gain so half input volume from perfect wave..DAC. And the passive SC-S2 came out to play tonight in front firing mode. An nht X1 of course dialed , PC -19 Dh440 monoblock doing duty on less than 70khz. Dunlavy. The man was one of the brilliants. Living the DAL legacy. That was PS ? Peace love and hi fidelity..
The 0R47 resistors are good to have as a lot of pass projects use them. Arrow had them priced recently at 9 cents a piece so I grabbed 300 of them.
Are you intending to try using the amp carcasses for some first watt builds?
Are you intending to try using the amp carcasses for some first watt builds?
Crest audio carcass. 2mm steel. Antek toroids now have nice reasonably priced cases.
I have the same amp. Do these resistors go bad after some time & exposure to temperature ? recently replaced all the power caps but nothing else
if amp is not broken due to some other circumstances, these 3W MOX will outlive us