Tweet
From http://www.divdev.fsnet.co.uk/repair2b.htm:
I have the same resistor burned that this guy had. Replaced... burned another. Do I need a reverb transformer?
Danny
Peavey Valverb, 1994 valve / transistor hybrid
Valve Reverb, 1U case prob. 1994 , no output.
One obvious problem. Seems to be C-R-C-R-C HT ps with
the first R overheated to partially white body and O/C.
Scraped back and remnants measure 4K and 12 K. Comparing
with other MO 1 W Rs, band positions on bodies, the band heated to black
was
probably gold and the first 2 Brown then blue , third is gone totally, .
so maybe 16K , replaced with 2 x 0.5W 33K, not higher wattage .
Supplying HT to 3 valves 12AT7 and 12AX7 so common enough.
The second dropper in
the ps drop down chain is 22K and looks fine.
Apparently no separate secondary for heater voltage - seems to use the
same LV winding used for the transitor power rails.
Wax coloured hot-melt glue, not wax under caps which looked fine
as well as all else.
What I thought was just an intermediary smoothing stage is actually also a
feed to a 12AT7 via a presumably impedance matching output transformer
to A2 anode and the other via 47R to the A1.
This transformer unpowered resistance seems ok about 2K on high imp
side and 300 on low imp which directly connects to one of the coil
type transducers on the spring-line. So maybe excess current in this valve
hopefully rather than a failing impedance matching transformer.
I've only ever come across piezo transducers on spring-lines before
this is 56R on the send? transformer connected side and 196R on receive end
coils.
I replaced with same wattage R ,disconnected valves and cut trace to
matching transformer.
Powered up via variac to full, giving unloaded HT rails of 440V or so
sustained with no problems.
On load hopefully the 440V should drop to less than design
max of 300V.
Dug out my old Avo 160 valve tester , noticed
last
calibrated in 1984 for the REME in the Hebrides. Also powered up via
variac initially. All those thumbwheel switch contacts and other contacts
that
could easily corrode but no problems.
I'd forgotten about the sequence , balance mains V, heater check, cold
insulation then hot
insulation , Test and Gasified check. Also forgotten that nifty 'telephone
dialler' for set mA/V.
All 3 valves nicely in the green GOOD sector. For the 12AT7 design
value is 5.5 mA/V and on test is 5.4 and 5.9 for the other triode.
Ignoring the drop associated with the other 2 valves and not
at the moment knowing the actual anode current, if at the test
value of 10mA then drop over 16K would give a dissipation of
1.6W. I cannot find in the data book or tester manual whether
that is 10mA each triode or combined value but I assume it
is per triode and 20mA if both sections on at test levels so 3.2W.
Didn't think to check for primary to secondary short on the matching
transformer. This was the problem , the bobbin is made of soft plastic
and with time/temeperature the ends had flexed enough to allow
some of the outer, LV, turns to drop into the HV turns area.
Removing the transformer and deforming the bulk of the colis
showed changing bridging resistance. Resistance of LV coil 310 ohm
and 2100 ohm for HV. Weight excluding bobbin about 40gm.
Average 'circumference' of LV coil 83mm and HV 58mm.
Wire 4 thou or 42swg so resistivity per 1000 yards 1910 ohm and
weight 0.145 lbs. So by weight predicted number of turns 1150
and by resistance 1790 turns. Warming up the coils with low set
hot air gun and removing vinyl tape and waxed paper and mounting
on coil winding machine gave counted-off number of turns 1800.
Before rewinding adding some vinyl tape either side of the
core and against the former, so that this coil doesn't slip
into the HV coil through the gap.
As HV section left undisturbed then implied by resistance then
about 17,500 turns for the HV section.
Representative voltages , no signal, controls at minimum
HTs 410V,246V,194V
At fuses 20V,20V ac
On 22uF electrolytics 0.6,0.6V and 4.6,17.6V dc
On the 3 valves DC
142,0,1,0,0,125,0,1,5.9
136,0,0.8,11.6,11.6,192,13,36,5.6
232,0,3,17.5,17.6,232,0,3,117
Also disabled the 240V/110V switch for use in UK - fiddlers
moving switch setting in USA would do no harm but in UK
a different matter. Some point in the future stops working
so lets turn/move every exposed switch and knob.
Valve Reverb, 1U case prob. 1994 , no output.
One obvious problem. Seems to be C-R-C-R-C HT ps with
the first R overheated to partially white body and O/C.
Scraped back and remnants measure 4K and 12 K. Comparing
with other MO 1 W Rs, band positions on bodies, the band heated to black
was
probably gold and the first 2 Brown then blue , third is gone totally, .
so maybe 16K , replaced with 2 x 0.5W 33K, not higher wattage .
Supplying HT to 3 valves 12AT7 and 12AX7 so common enough.
The second dropper in
the ps drop down chain is 22K and looks fine.
Apparently no separate secondary for heater voltage - seems to use the
same LV winding used for the transitor power rails.
Wax coloured hot-melt glue, not wax under caps which looked fine
as well as all else.
What I thought was just an intermediary smoothing stage is actually also a
feed to a 12AT7 via a presumably impedance matching output transformer
to A2 anode and the other via 47R to the A1.
This transformer unpowered resistance seems ok about 2K on high imp
side and 300 on low imp which directly connects to one of the coil
type transducers on the spring-line. So maybe excess current in this valve
hopefully rather than a failing impedance matching transformer.
I've only ever come across piezo transducers on spring-lines before
this is 56R on the send? transformer connected side and 196R on receive end
coils.
I replaced with same wattage R ,disconnected valves and cut trace to
matching transformer.
Powered up via variac to full, giving unloaded HT rails of 440V or so
sustained with no problems.
On load hopefully the 440V should drop to less than design
max of 300V.
Dug out my old Avo 160 valve tester , noticed
last
calibrated in 1984 for the REME in the Hebrides. Also powered up via
variac initially. All those thumbwheel switch contacts and other contacts
that
could easily corrode but no problems.
I'd forgotten about the sequence , balance mains V, heater check, cold
insulation then hot
insulation , Test and Gasified check. Also forgotten that nifty 'telephone
dialler' for set mA/V.
All 3 valves nicely in the green GOOD sector. For the 12AT7 design
value is 5.5 mA/V and on test is 5.4 and 5.9 for the other triode.
Ignoring the drop associated with the other 2 valves and not
at the moment knowing the actual anode current, if at the test
value of 10mA then drop over 16K would give a dissipation of
1.6W. I cannot find in the data book or tester manual whether
that is 10mA each triode or combined value but I assume it
is per triode and 20mA if both sections on at test levels so 3.2W.
Didn't think to check for primary to secondary short on the matching
transformer. This was the problem , the bobbin is made of soft plastic
and with time/temeperature the ends had flexed enough to allow
some of the outer, LV, turns to drop into the HV turns area.
Removing the transformer and deforming the bulk of the colis
showed changing bridging resistance. Resistance of LV coil 310 ohm
and 2100 ohm for HV. Weight excluding bobbin about 40gm.
Average 'circumference' of LV coil 83mm and HV 58mm.
Wire 4 thou or 42swg so resistivity per 1000 yards 1910 ohm and
weight 0.145 lbs. So by weight predicted number of turns 1150
and by resistance 1790 turns. Warming up the coils with low set
hot air gun and removing vinyl tape and waxed paper and mounting
on coil winding machine gave counted-off number of turns 1800.
Before rewinding adding some vinyl tape either side of the
core and against the former, so that this coil doesn't slip
into the HV coil through the gap.
As HV section left undisturbed then implied by resistance then
about 17,500 turns for the HV section.
Representative voltages , no signal, controls at minimum
HTs 410V,246V,194V
At fuses 20V,20V ac
On 22uF electrolytics 0.6,0.6V and 4.6,17.6V dc
On the 3 valves DC
142,0,1,0,0,125,0,1,5.9
136,0,0.8,11.6,11.6,192,13,36,5.6
232,0,3,17.5,17.6,232,0,3,117
Also disabled the 240V/110V switch for use in UK - fiddlers
moving switch setting in USA would do no harm but in UK
a different matter. Some point in the future stops working
so lets turn/move every exposed switch and knob.
Danny
Comment