• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Design of C-core nanocrystalline GM-70 output transformer

In practice you can have some air clearance between the bobbin and core that will take off some capacitance. It is especially important for nanocrystalline and amorphous C-cores, most of them having a distorted shape. I usually leave my cores floating for OPTs.
 
I wouldn't standardize if primary or secondary as a start layer is better. Both have advantages and disadvantages. If you start with a primary package, you get the advantage of less capacitance between the secondary. But you can do the same with a secondary starting transformer by inversing the winding of an internal primary package. Etc, etc. I'm using both strategies, depending of my transformer model.
What I like in secondary starting transformer is by respecting the 1/2 outer layer rule, I make the outer secondary layers 1R equivalent and the inner layers 4R equivalent, which gives me an easier Williamson 4R, 9R and 16R connection style.
Whereas in primary starting transformer, where I have to respect all secondary layers being equal, I have to intermix the 1R and 4R layers together, either by going trifilar or different windings on top.
 
Bac > Bdc means added distortion even if it is not saturating. There is no zero crossing in SE amps. It is a good rule to keep Bdc a bit higher than Bac full power at the lowest frequency of interest. I suggest 30Hz.
 
With a higher Bac, you basically get 2nd harmonic in the upper north-east direction of the quadrant (lower flux density headroom), and zero crossing into the bottom south-west direction. Ideally, you want to balance the two 50%-50% for a chosen bottom frequency, like @45 suggested.
 
I would be worried more about the high level of overall distortion rather harmonics. Until Bac becomes equal or smaller (at significantly higher frequency in your example) it is likely (almost certain) you will get double digit distortion ( up to 15-20% or more could be expected). Which harmonics is really not relevant. No free lunch.
 
high level of overall distortion rather harmonics. Until Bac becomes equal or smaller (at significantly higher frequency in your example) it is likely (almost certain) you will get double digit distortion ( up to 15-20% or more could be expected). Which harmonics is really not relevant.
I'm sorry but I don't get it. In the situation I described (Bdc=0.3T, Bac=0.8T, Bsat=1.2T) I'm not reaching core saturation ( -0.5T < B < 1.1T ) so what is the reason of this high distortion?
 
You will reach higher distortion, compared to 0.6 to 0.6 Bac to Bdc.
Considering Bsat of 1.2 is on the extreme high for nanocrystalline cores. The knee for ungapped regular nanocrystalline C-cores begins at 0.8T. I consider 1T to be maximum.
Try tracing a magnetizing curve for your cores to check this out. You could take off the curves using a variac, series resistor, Vac source and some voltmeters.
 
Ok to left behind distortion issue and back to my uncertainty lets take Bdc=0.3T and Bac=0.5T => -0.3T < B < 0.8T.
Is it OK for SE OPT to allow zero crossing?
I'm asking because potentially I'll move to GMI-30 tube (Rinternal=3.8k Ranode=15k Ia=100mA) and its hard to find sweet spot.
 
I'm sorry but I don't get it. In the situation I described (Bdc=0.3T, Bac=0.8T, Bsat=1.2T) I'm not reaching core saturation ( -0.5T < B < 1.1T ) so what is the reason of this high distortion?
Because your zero is Bdc and hysteresis (here a straight line with very good approx., thanks to air-gap) develops symmetrically around it. In a SE OPT you only have one quadrant to work with. So negative values are not possible and you will get an anomalous behaviour at the origin (no zero-crossing). Your hysteresis will be "squashed" on the opposite side of saturation and that generates quite some distortion too. It will become symmetric again when Bac is smaller than Bdc which, for this transformer, can only happen increasing frequency.
I wish it were so simple because SE output transformers could be much smaller for a given application.

If your Bmax= Bdc + Bac is 1T, then Bdc for 100 mA should be no less than 0.5T, better if a bit higher.
 
Last edited:
Is it good idea to move some layers from outer primary sections to inner ones to balance C and R of this sections?

eg:
from:
CORE - 4L (P) - 1L (S) - 4L (P) - 1L (S) - 4L (P) - 1L (S) - 4L (P)
to:
CORE - 5L (P) - 1L (S) - 5L (P) - 1L (S) - 3L (P) - 1L (S) - 3L (P)
 
At first glance, I would rearrange the primary to start and end with 3 primary layers.
If space and other consideration allow you migth consider
Core -2L(P) - 1L(S) - 4L(P) - 1L(S) - 4L(P) - 1L(S) - 4L(P) - 1L(S) - 2P(P).
Starting and ending with half of a midsections ampere windings would reduce leakage inductance to 1/4 .
At the same time it would offcourse increase P-S capacitance wich, if possible, could be counteracted by increasing P-S isolation.
8 P-S isolation layers instead of 6 would offcourse increase P-S capacitance and/or total winding heigth wich may be unacceptable.

If i, for whatever reason, would have to go with 3 secundary layers i still would always try to flank every secondary with half of the primary winding each side first, and only abond it if p-s capacitance and/or resulting undesirable Q dictates otherwise.
My first choice:
half (section) primary amperewindings - full (section) secondary Aw - full (section) primary Aw ...a.s.o.. ending in half a sections amperewindings.
 
Last edited:
  • Like
Reactions: Marek
Some GM70 spice models and optional static setting points. With 12K static load line.
(it is not measured chrs. but from PDF from internet)
DHT and classic PSpice models.

Code:
**** GM70 ******************************************
* Created on 06/16/2024 11:04 using paint_kit.jar 3.1
* www.dmitrynizh.com/tubeparams_image.htm
* Plate Curves image file: /Users/zoran/Desktop/Model_Paint_Tools/Graph/GM70.jpg
* Data source link:
*----------------------------------------------------------------------------------
.SUBCKT TRIODE_GM70 1 2 3 ; Plate Grid Cathode
+ PARAMS: CCG=8P  CGP=12P CCP=4P RGI=2000
+ MU=8.09 KG1=2760 KP=170 KVB=3000 VCT=4 EX=1.415
* Vp_MAX=2200 Ip_MAX=300 Vg_step=20 Vg_start=40 Vg_count=16
* Rp=12000 Vg_ac=110 P_max=125 Vg_qui=-110 Vp_qui=1120
* X_MIN=66 Y_MIN=78 X_SIZE=966 Y_SIZE=615 FSZ_X=1867 FSZ_Y=800 XYGrid=true
* showLoadLine=y showIp=y isDHT=n isPP=n isAsymPP=n showDissipLimit=y
* showIg1=n gridLevel2=n isInputSnapped=n 
* XYProjections=y harmonicPlot=y dissipPlot=y
*----------------------------------------------------------------------------------
E1 7 0 VALUE={V(1,3)/KP*LOG(1+EXP(KP*(1/MU+(VCT+V(2,3))/SQRT(KVB+V(1,3)*V(1,3)))))}
RE1 7 0 1G  ; TO AVOID FLOATING NODES
G1 1 3 VALUE={(PWR(V(7),EX)+PWRS(V(7),EX))/KG1}
RCP 1 3 1G   ; TO AVOID FLOATING NODES
C1 2 3 {CCG} ; CATHODE-GRID
C2 2 1 {CGP} ; GRID=PLATE
C3 1 3 {CCP} ; CATHODE-PLATE
D3 5 3 DX ; POSITIVE GRID CURRENT
R1 2 5 {RGI} ; POSITIVE GRID CURRENT
.MODEL DX D(IS=1N RS=1 CJO=10PF TT=1N)
.ENDS
*$

**** GM70 Composite DHT *****************************************
* Created on 06/16/2024 11:04 using paint_kit.jar 3.1
* www.dmitrynizh.com/tubeparams_image.htm
* Plate Curves image file: /Users/zoran/Desktop/Model_Paint_Tools/Graph/GM70.jpg
* Data source link:
*----------------------------------------------------------------------------------
.SUBCKT DHT_GM70 1 2 3 4 ; P G K1 K2
+ PARAMS: CCG=8P  CGP=12P CCP=4P RFIL=6.667
+ MU=8.09 KG1=2760 KP=170 KVB=3000 VCT=4 EX=1.415 RGI=2000
* Vp_MAX=2200 Ip_MAX=300 Vg_step=20 Vg_start=40 Vg_count=16
* Rp=12000 Vg_ac=110 P_max=125 Vg_qui=-110 Vp_qui=1120
* X_MIN=66 Y_MIN=78 X_SIZE=966 Y_SIZE=615 FSZ_X=1867 FSZ_Y=800 XYGrid=true
* showLoadLine=y showIp=y isDHT=y isPP=n isAsymPP=n showDissipLimit=y
* showIg1=n gridLevel2=n isInputSnapped=n 
* XYProjections=y harmonicPlot=y dissipPlot=y
*----------------------------------------------------------------------------------
RFIL_LEFT    3  31 {RFIL/4}
RFIL_RIGHT   4  41 {RFIL/4}
RFIL_MIDDLE1 31 34 {RFIL/4}
RFIL_MIDDLE2 34 41 {RFIL/4}
E11 32 0  VALUE={V(1,31)/KP*LOG(1+EXP(KP*(1/MU+V(2,31)/SQRT(KVB+V(1,31)*V(1,31)))))}
E12 42 0  VALUE={V(1,41)/KP*LOG(1+EXP(KP*(1/MU+V(2,41)/SQRT(KVB+V(1,41)*V(1,41)))))}
RE11 32 0 1G
RE12 42 0 1G
G11 1 31 VALUE={(PWR(V(32),EX)+PWRS(V(32),EX))/(2*KG1)}
G12 1 41 VALUE={(PWR(V(42),EX)+PWRS(V(42),EX))/(2*KG1)}
RCP1 1 34 1G
C1 2 34 {CCG} ; CATHODE-GRID
C2 2 1 {CGP} ; GRID=PLATE
C3 1 34 {CCP} ; CATHODE-PLATE
D3 5 3 DX ; FOR GRID CURRENT
D4 6 4 DX ; FOR GRID CURRENT
RG1 2 5 {2*RGI} ; FOR GRID CURRENT
RG2 2 6 {2*RGI} ; FOR GRID CURRENT
.MODEL DX D(IS=1N RS=1 CJO=10PF TT=1N)
.ENDS
*$
.
scroll down to fig17.
It is important - longer life notes. Without loosing much output power...
https://www.turneraudio.com.au/loadmatch-1-SE-triodes.html
.
also some infos here:
same anode chrs
https://www.bartola.co.uk/valves/2018/03/13/gm-70-se-amplifier/
.
another page with gm70:
https://www.audiodesignguide.com/Claudio845/813amp.html
.

G70 longer life settings 12K.png

G70 longer life settings 14K.png