Knoppson OPT Test Results

- KT66 push-pull drive is intended if otherwise not noted (2rp=2.5k Ohm, Zl=10ka-a@8Rload)
rp for KT88/6550/300B/PX25 equals approximately half that for KT66 (2rp=1.25k)
For test method description and discussion, see OPT Dynamic & Static Performance Testing

Test method





1.1) Lp(5V, 10, X)

(84, 91, 88)H

(88, 99, 91)H

(270, 322, 1028)H

X=(459, 462, 460)V

1.2) fl (KT66, KT88)

(3.8, 2.1)Hz

(3.6, 2.0)Hz

(1.2, 0.65)Hz


2.1) Phi(f1, f2, f3)

(15, 26, 27)Deg

(45, 51, 62)Deg



2.1.1) Lleak/Ci

(47, 34, 47, 43)mH

(6.0, 5.3, 6.2, 6.4)nF

(28, -, -, -)mH

Ci for rev. D

2.1.2) wLleak’

(496, 288, 288)Ohm

(501, 299, 299)Ohm

(-, 176@1kHz, -)Ohm

dLp=(0, 0, 0)mH

2.2) fh(KT66, KT88)

(42, 38)kHz

(17, 23) kHz

(71, 64) kHz

Verified for rev. D

3) Rcu(prim,sec)

(209, 0.26)Ohm

(206, 0.26)Ohm

(180, 0.17)Ohm


5) L(In, Imax, Iwc)

(73/4, 57/7, 29/27)

(60/4, 57/7, 28/26)

(19/4, 9/7, 2.5/26)

[H/mAdc], dI(A, 50H)=1.7

6) Step(Trise, f, %)

(8.0, 30, 1.6)

(16, 0, 0)

(10, 62, 2.0)

[us, kHz, %]

1.1) Lp is evaluated at (5V/50Hz, 10V/50Hz, XV/50Hz), where X is noted and tested as a transformed secondary
n^2*Ls evaluation at approximately twice its normal operating voltage or 460V/50Hz=230V/25Hz. Lp(5V/50Hz)
should by original Williamson recommendation exceed 95H for fl<3.3Hz (for BF+1=20dB).

1.2) wl*Lp=2rp//Zla-a (=2k). fl is evaluated at Lp(5V/50Hz) only. Observe that low-end frequency response will
improve for lower plate or load impedances. First value is calculated for KT66/8R, second for KT88/8R.

2.1.1) Lleak is evaluated at 5V/(MEAN, 15kHz, 20kHz, 23kHz) with the use of the formula
Tan(Phi)=wLleak/(Zla-a+2rp) or wCi(Zla-a//2rp) depending on Phi(Zdrive) behaviour. Lleak should by original
Williamson recommendation never exceed 35mH for fh>60kHz (for BF+1=20dB).

2.1.2) wLleak’ is evaluated at 5V/50Hz and (phase, sc; anti-phase, sc; anti-phase, oc):
Z(Phase, sc)=jwLleak’+Rcu, wLleak’^2=|Z|^2-Rcu^2
Z(Anti-phase, sc)=jwLleak’+Rcu, wLleak’^2=|Z|^2-Rcu^2
Z(Anti-phase, oc)=jw(Lleak’+dLp)+Rcu, wLleak’’^2=|Z|^2-Rcu^2
Lleak’(Phase, sc) should equal Lleak’(Anti-phase, sc) if and only if HF loss model holds. X(Lleak)<0.05Rcu.

2.2) wh*Lleak=Zla-a+2rp (=12.5k). fh is evaluated at Lleak(MEAN) only. Observe that high-end frequency response
will improve for higher plate or load impedances (as long as Ci doesn’t become significant). First value is calculated
for KT66/8R, second for KT88/8R.

3) Rcu(prim) is preferably done at anti-phase connected primaries (eliminating DVM reading oscillations). Rcu(sek)
may most easily be estimated by connecting a 10R/1% resistor in series with the secondary and a 1.5V battery.

5) Connect a 1.5V battery in series with a 330R resistor in parallel to the OPT primary windings as well as the
(>>200Ohm Reostat regulated) driving ac-voltage source. Tune the ac-source to 5V/50Hz over the whole primary.
Measure Iac, Idc and Vdc (which will be around 0.5Vdc which also is the recommended normal plate voltage offset).
Double Idc for dI. Calculate L(dI). Switch to a 100R resistor and remeasure L(I(1V)). Finally, switch to a 9V battery
and reuse the 330 Ohm resistor for the worst-case L(I(3V)).

6) Apply a 5V/10KHz square wave thru the 2.5k resistor with 8R at the secondary. Measure at the secondary.
Note Trise, any oscialltion frequency and its percentage amplitude (<4% for fh=0.35/Trise).