[DSP-10] Brickette QRO QRV at n5bf/6

Sun Jul 3 19:45:44 EDT 2005

Perry,

This hint is a lifesaver!  It means that Figure 1 and Figure 2 in the 
article agree and that "looks to me like" was plus one.  L2, L4, and L5 
are now all returned to and retuned at five turns and I have learned how 
to look at or wind a toroid.

L3 remains at five turns and is retuned to give a good SWR indication 
across the band.  After other cleanup I'm now getting 5.7 watts out peak 
for a DSP-10 drive of 95.  (Going up to 100 doesn't increase output, as 
expected.)

After studying the datasheet on the module again, I decided it was worth 
checking if increasing the Vcc on pins 2 and 4 would help.  D8 is the 
big drop but there are little ones on the fuse and FBs.  Increasing Vcc 
from 12.4 to 12.6 improved output by 0.1 - 0.2.  If I fixed up all my 
supply wiring to get 13.2 or so on the part, maybe output would get into 
the 7 W range.  Still not 8 or 9.  Spec. (well, "typ.") is >= 7 watts @ 
12.5 Vcc so I'm still not getting near that at 12.6.

Having looked at just about everything else on the schematic, on the 
datasheet, and in the box, I think there are only two choices at this point:

1.  Live with it, characterize it, and call this done.
2.  Order a new module, install it, characterize it see if it is 
significantly different (hopefully better).

It being a holiday tomorrow, it wouldn't make any difference whether I 
placed an order today or tomorrow, so I think I'll sleep on it.

Thanks again for everyone's helpful advice.

Courtney

Perry Ogletree wrote:
> Hint... When winding toroid cores, each time the wire passes through the 
> center is a "turn", one turn would be a simple "horseshoe" of wire, two 
> turns would be a single loop, etc.
> 
> 73 de Perry - K4PWO
> 
> ----- Original Message ----- From: "Courtney Duncan" 
> <cbduncan at earthlink.net>
> To: "kd7ts" <kd7ts at ispwest.com>
> Cc: "Discussion of DSP-10 2-meter transceiver" <dsp-10 at mailman.qth.net>
> Sent: Saturday, July 02, 2005 6:38 PM
> Subject: Re: [DSP-10] Brickette QRO QRV at n5bf/6
> 
> 
>> Yes, I have a copy of the QST article, in fact, a color one that I 
>> printed from arrl.org.  Higher resolution pictures!
>>
>>
>> kd7ts wrote:
>>
>>> Confirm the winding of L4 and L5.
>>
>>
>> This is a little confuing.  Figure 1 says 5 turns on a T-25-17, but in 
>> Figure 2 it looks like four turns to me.  I have tried both what looks 
>> to me like four turns and five turns.
>>
>> Originally I built it with what looks to me like 5 turns on l2, l3, 
>> l4, and L5.  To get it working with the SWR light out, I removed one 
>> turn from each.  At that point I had inductors that matched Figure 2 
>> and 5.1 watts out forward.
>>
>>>
>>> If these are wrong, spreading or compressing their turns has very
>>> little affect on tuning.
>>>
>>
>> In that configuration, spreading or compressing L4 or L5 had little 
>> effect.  Today I took them out and rewound them with what looks to me 
>> like 5 turns again.  Now I can get some roll off by compressing the 
>> turns.  If I push them all up together, it drops under three watts.  I 
>> think this is the right behavior.  Increasing the inductance like that 
>> should lower the cutoff frequency of the low pass filter.  I spread 
>> them out to where I'm getting 4.9 watts forward again.  On the 
>> schematic it looks like L2 should be involved in low pass filtering 
>> also, so I took it out and rewound it too to what looks like to me is 
>> five turns.
>>
>> With this I can get reduced power by compressing L2 also and so spread 
>> it out to where it does 4.9 watts forward again.
>>
>> Since this is a low pass filter and since I possibly had inductors too 
>> small, it is possible that the difference between 4.9 and 5.1 watts 
>> forward on the Bird wattmeter to a 50 ohm resistive dummy load is 
>> power in a harmonic that is now adequately suppressed.  I don't have 
>> any way here to measure for this hypothesis but, thinking this, I 
>> think I'll leave it this way.
>>
>> I cannot see how L3 is involved with anything but turning the RF and 
>> SWR lights on so I left it as it was, at what looks to me like 4 
>> turns.  It had to be retuned to re-extinguish the SWR light and now 
>> I'm getting expected RF and SWR light behavior again.  Is there some 
>> reason why L2 and L3 have to be the same value?  I simply couldn't get 
>> reflected power to null with L3 at what looks to me like 5 turns.
>>
>>> If the peak power (forward) is well defined by changing the spacing,
>>> they are correct.
>>>
>>
>> The peak should be broad, right?  I'm getting a peak at about the 
>> widest spacings on L4 and L5 (and L2) and can certainly reduce power 
>> by compressing turns now.
>>
>>> check the pins of the T/R relay for connection and solder bridges.
>>
>>
>> Done.  Nothing anomalous found.
>>
>>>
>>> confirm the input and output coax shields are connected at both ends.
>>>
>>
>> I originally used a good piece of some scrap RG-58 for this which was 
>> big, ugly, hard to bend, and could have been a problem.  I've replaced 
>> all three RF connector jumpers (Drive, Rx, Ant) with nice miniature 
>> coax.  The ends are all pretty and well made.  There is no change in 
>> max output.
>>
>>> Do you have a copy of the original QST article ?
>>>
>>> 73 KD7TS
>>>
>>> Mike
>>>
>>>
>>>
>>
>> So, with all this, I still think I just have an underperforming 
>> module. The only other thing I can think of is inadequate DC but I 
>> don't time right now to try to get above the 12.4 V (after connector 
>> drops and things from a 13.8 V regulated supply) that I currently 
>> have, when at max out (13.6 Rx, and I think I could rewire the power 
>> to get like 13.0 at max out).
>>
>> The low pass filter is supposed to be a matched low pass filter, not 
>> particularly a matching network, right?  Everything should be 50 ohms 
>> end to end and the low pass filter is just designed to be that way but 
>> not to deal with any mismatch.  Yes?
>>
>> Oh, on the matching network L1 and C4, I do see a few percent change 
>> when I bend L1 around.  I have it peaked right now too.
>>
>> Thanks again Mike for the consultation.  This isn't where I can't live 
>> with it operationally, but the mystery is part of the fun and if there 
>> is something grossly wrong I'd like to understand and fix it.
>>
>> Courtney
>> _______________________________________________
>> DSP-10 mailing list
>> DSP-10 at mailman.qth.net
>> http://mailman.qth.net/mailman/listinfo/dsp-10
>>
> 
> 
> _______________________________________________
> DSP-10 mailing list
> DSP-10 at mailman.qth.net
> http://mailman.qth.net/mailman/listinfo/dsp-10
> 