My Ampeg SVT preamplifier build : documentation available

[PRR]

Jon Fleig's Ampeg SVT (SVP-CL) preamp voltage chart from LTSpice Sim B+ = 277 VDC
------ --------- ---------  Plate Cath Grid
V1A 12AX7 Gain #1 --  150 1.53 0
V1B PostEQ CF -------- 255 91 0.019  ----------- ***
V2A 12AU7 ULow CF -- 255 22 150 direct coupled to V1A
V2B UL TS recovery --  152 29 0  ----------- ***
V3A 12AX7 James TS - 145 18 17
V3B Midrange drive --- 103 18 0  ----------- ***
V4A 12AU7 Mid CF ---- 277 110 103 direct coupled to V3B
V4B Output CF -------- 277 6.4 152   ----------- ***

*** sure do not look like I would expect on a meter. Are you maybe citing G-K and P-K voltages, rather than to-ground voltages?

[vibrolax]

Jon Fleig's Ampeg SVT (SVP-CL) preamp voltage chart from LTSpice Sim B+ = 277 VDC
------ --------- ---------  Plate Cath Grid
V1A 12AX7 Gain #1 --  150 1.53 0
V1B PostEQ CF -------- 255 90 91  ----------- ***
V2A 12AU7 ULow CF -- 255 153 150 direct coupled to V1A
V2B UL TS recovery --  152 7.25 0  ----------- ***
V3A 12AX7 James TS - 104 10.6 9.9
V3B Midrange drive --- 223 10.6 0  ----------- ***
V4A 12AU7 Mid CF ---- 277 222 223 direct coupled to V3B
V4B Output CF -------- 277 157 152   ----------- ***

*** sure do not look like I would expect on a meter. Are you maybe citing G-K and P-K voltages, rather than to-ground voltages?

All the voltages were obtained from the LTSpice probe tool, which AFAIK, shows ground-ref voltages.
I ran the sim again this morning (not making any changes), but obtained different results.  I have updated these in PRR's table (updated lines are green). I cannot explain why the results were different.  Either I made some mistake, or LTSpice computed the DC operating points differently.  I may re-do the editing/sim process starting from the original file to see if I can reproduce the behavior.

[vibrolax]

I can reproduce the behavior that different DC operating points are shown for supposedly identical sim conditions, depending on the run state of LTSpice IV and LTSpice XVII.
I obtain significantly different results for a few nodes between XVII and IV.  It will take longer to understand why these behaviors occur.  I also need to test if I observe this behavior on other models I have created.

Yes, this is an example of the pitfalls of believing model results without checking against reality and/or approximate independent calculations.

[PRR]

OK, I was not sure what the schematic was. Your "Documentation" page says schematic but is really a layout. Drilling down I find a "schematic" which is a node-list and not meant for human consumption. I assume the image in Reply #51 is the best current schematic?

> approximate independent calculations

Single tube stages can usually be computed on sight. Grid-cathode voltage tends to be *about* plate-cathode voltage divided by Mu. At U3 we have Mu=100, Vp=150V, Vg=0V, Vk=1.5V, check. At U5 we have Mu=20, Vp-k=100V, Vg-k=4V, check (not concerned 4V or 5V).

U8 is a special case because, as drawn, there is heavy DC path from U9 K back to U8 K. This gives 10Vgk on U8, and Mu=100 times 8V suggests 800V at plate, which can't happen. So U8 is near cut-off, as shown by very high Vp (and low Vpk on U9).