[JMS] Power supply design concerns

Sat, 21 Feb 2004 00:08:42 -0600

This ain't rocket science.  The solution is simple.  Keep the 600V 
transformer in the design.  Make the filter a choke input by moving the 
input capacitor to the output to reduce the voltage to .9 x 300 = 270VDC. 
270 x 100 ma = 27 Watts DC input.  At 66% efficiency you'll have 17 Watts 
output (perfect).  And, your modulator current is not going to be 140 mA 
all the time. If it peaks at 140 mA, the average will be around 1/3 that, 
or 45 mA.  45+100+40+35 = 220 mA.  Your choke will be fine.  If you don't 
believe me, lash the system up with clip leads to prove it works before 
you start drilling holes.  Having said that, it is always better to have 
separate power supplies for the RF Amp and the Modulator so if you have 
the extra transformer then use it.

Brian K. Harris, Account Manager

Philips Semiconductors 
2140 Lake Park Blvd. Suite 200
Richardson, Texas 75080 USA 

Office:     972-705-2484 (Use This First)
Fax:         972-705-2450
Mobile:    214-763-5977
EMAIL [email protected]

[email protected]
Sent by: 
[email protected]
2004-02-20 11:47 AM
Please respond to james_millen_society

        To:     [email protected]
        cc:     (bcc: Brian K Harris/SVL/SC/PHILIPS)
        Subject:        [JMS] Power supply design concerns
        Classification: 

All:
Just wanted to keep you abrest of the situation with regards to my Millen 
transmitter project. Yesterday, a 10.5" by 19" steel rack panel came in 
(which will be the basis for my power supply chassis). I ran into some 
design problems along the way. Thinking that I can supply both the 
modulator and transmitter, I started on paper with my 600 VCT 250 mA 
Hammond transformer and a 12H/240 mA choke. 

The minimal bleeder value to use the 12H choke is 35 mA (if you use the 
formula B+/L to give the current in mA). So 420 VDC (that's 300 * 1.414) 
div by 12 H. I think this is to ensure that the choke will already be in 
its operational curve to ensure regulation. On CW, this will be important. 
I did build into the circuit a voltage divider so to maintain a minimum of 
35 mA, a 12kohm resistor at 30 watts should be used. That meant that I 
could use a string of two 6kohm resistors to split the 420 VDC down to 210 
VDC (ideal for the modulator screen voltage). 

Unfortunately, I started to add up the current draw needed to power the 
transmitter AND the modulator: 35 mA for the bleeder, about 40 mA for the 
mod screen, probably about 140 mA for the modulator plate (if a minimum of 
55 Watts is needed for cathode modulating the 90801, and another 90-100 mA 
for the transmitter. That is well beyond the 240 mA limit for the choke! 
Not only that, if the supply gives the transmitter around 400 VDC for the 
807 plate, were talking about a 40 Watt power input. The 90800 transmitter 
seems to exhibit a 66% efficiency so that translates to 26 watts out to 
the 812A's. You need only 17.5 watts to drive the 812A's according to the 
Millen data sheets. Already, it looked bad. Luckily, I had another 
transformer lying around that would take away some of the "pressure" on 
the supply. I have a 375 VA/205 VAC (no c.t.) transformer lying around. 
Instead of loading the power supply with both the modulator and the 
transmitter, I decided to power the plate suppl
 y of the 90800 with this power supply setup. Besides, power input to the 
812A's comes out to be around 19 watts, much closer to the spec of the 
Millen data sheet! Well, now the modulator supply seems to be fine with 35 
mA at the bleeder, 40 mA at the screens of the 6146's, and 140 at the 
plate. That's 215 mA total current. I have a little 25 mA headroom but to 
be quite honest, I don't think I'll be using the full 215 mA on this 
supply. So, that's the situation with the power supply. Looking forward to 
your thoughts. Cal, N6KYR.

________________________________________________________________
The best thing to hit the Internet in years - NetZero HiSpeed!
Surf the Web up to FIVE TIMES FASTER!
Only $14.95/ month -visit www.netzero.com to sign up today!
______________________________________________________________
James_Millen_Society mailing list
Home: http://mailman.qth.net/mailman/listinfo/james_millen_society
Help: http://mailman.qth.net/faq.htm
Post: mailto:[email protected]