[Elecraft] XV50 PA Bias

[Dave Sergeant]

An update on my XV50.

Having failed to get the pa bias greater than around 500mA I decided to 
press ahead with the rest of the alignment. It is now basically working 
but only at the 5-10W level. If I increase the drive to 20W it 
effectively limits at that level, no more than 20W can be obtained and 
it is VERY flat topped on SSB. I assume this is because the pa is 
non-linear because of the incorrect bias. But it appears reasonably 
linear with the input attenuation R22 turned down to the lower power.

I also did the mod to my KAT2, changing the 47 ohm to 470 ohm - the 
power control on the K2 now works correctly.

I intend to leave it set at the 5-10W level and get some reports from 
locals - that is the sort of power I intend to use anyway.

I see there have been various similar postings on the reflector over 
the past few years with no real answers forthcoming. Looking at the 
datasheet of the RA30H0608M it is only specified down to 60MHz and the 
performance at 50MHz is totally unspecified. I suspect I, and others 
who have had these sort of issues, have devices at the edge of the spec 
and the only cure is to try a replacement, which really is pot luck.

73 Dave G3YMC

On 24 Oct 2014 at 13:49, Dave Sergeant wrote:

> R39 is correct (10k pot) as is R10 (100k) and all the measurements
> confirm this. The adjustment range, measured at pin 2 of the amplifier
> module, is 2.5-4.7V, also what is expected from the circuit. The maximum
> voltage is limited to 4.7V because of the 5mA flowing through R28 which
> swamps the 1mA bias current.
> 
> What I would like to know is what the bias voltage should be on a 
> correctly set up XV50. The datasheet seems to indicate it must be at
> least greater than 4.5V. It seems my amp needs more than 4.7V to get the
> correct bias, and judging from previous posts on this subject I don't
> think I am alone.
> 
> 73 Dave G3YMC
> 
> On 24 Oct 2014 at 7:46, Don Wilhelm wrote:
> 
> > Dave,
> > 
> > The PA bias control usually has much greater range than is needed, and
> > the PA bias input should draw very little current. Check the
> > resistance of R39.  Is there a chance that you swapped R39 and R10
> > during the assembly?
> > 
> > 73,
> > Don W3FPR
> > 
> > On 10/24/2014 6:46 AM, Dave Sergeant wrote:
> > > I am aligning my new XV50, latest build using the RA30H0608M module
> > > and having issues with setting the quiescent bias of the pa as per
> > > the manual. I see others have had this same problem but I can find
> > > no answers.
> > >
> > > The manaul calls for R39 to be adjusted for 20mV on the DMM across
> > > TP3/4 (across the 20mohm current sense resistor R17). With R39 at
> > > its maximum I get 16mV corresponding to a total current of 800mA as
> > > against the required 1A. This is confirmed by the digital current
> > > meter on the 12V psu powering it and my K2, which increases from
> > > 0.5A to 1.3A when I enable transmit on the XV50.
> > >
> > > Before anybody asks, I DO have link W1 fitted, and in fact I have
> > > the correct bias voltage on pin 2 of the output module. However
> > > there is where I suspect the problem as the maximum voltage
> > > achievable there is 4.7V.
> > >
> > > I found a datasheet of the RA30H0608M, albeit a preliminary one
> > > which has a few sheets 'in preparation'. It does not give any
> > > figures for standing bias versus the bias volts but suggests it must
> > > be somewhere between 4.5 and 5V (there is no power output below
> > > 4.5V). The voltage at the R29/R39 junction is limited by the current
> > > flowing through R28 and the module bias current (typically 1mA). So
> > > I am wondering whether some rf modules need a bit more than 4.7V to
> > > get the correct bias current of around 1A. Maybe reducing the value
> > > of R39 (56 ohms) is the answer. Or is mine near enough for good
> > > linearity?

http://davesergeant.com