PP logo


Transformer balanced line input amplifier.



IA19 is a 2 channel transformer input amplifier for audio.
The circuit was originally designed for Beyer Dynamics TR/BV 310 type transformers, but these are now out of production and I am out of transformers.
The description for the original circuit ( IA19A ) is available in the download section.
This revision ( IA19B ) can be used with Neutrik NTM 1 or Bourns LM-NP-1001 transformers.
Some of the design parameters are:


Audio transformers are not designed for and should never be used as mains-insulation transformers.
Even a transformer with an insulation rating of 6.5 kV ( like the Bourns type ) do not have sufficient insulation distances for use as mains-insulation transformers.

Transformer circuits.

Voltage transformer amplifier schematic.
Fig.1: Voltage transformer amplifier.

Standard voltage transformer amplifier.
This is the circuit that have the best performance at low source impedances, but for low THD at low frequencies the transformer is very large.
Lp and Ls are the primary and secondary winding inductances. Rp and Rs are the primary and secondary winding resistances.
Za is the transformer secondary load. This is normally specified in the transformer data sheet.
Ra and Rb are optional resistors to add gain.
See [1], page 27 for an example of circuit performance.

Current transformer amplifier schematic.
Fig.2: Current transformer amplifier.

This circuit shorts the output of the transformer as the input impedance into node A is close to 0. This will reduce the field in the transformer reducing the THD.
The gain of the circuit is R27 / ( R21 + R22 ) / n, where n is the transformer winding ratio.
The input impedance is R21 + R22.
For a small 200 Ω 1:1 transformer ( Beyer TR/BV 310 001 001, R21 = R22 = 4.99 kΩ, R27 = 10 kΩ ) this will increase the maximum input signal from -20 dBu to 21 dBu at 100 Hz and 0.02% THD.
This circuit is more noisy than the circuit of fig.1. A large part of the noise is from the 3 resistors R21, R22 and R27.
For an input impedance of 10 kΩ and a gain of 1 ( R21 = R22 = 5 kΩ and R27 = 10 Ω ), the noise contribution of these 3 resistors is -110 dBu.
Another issue is DC. A large part of the OP-AMP bias current is supplied through the transformer causing some 2. harmonic distortion.

AC coupled current transformer amplifier schematic.
Fig.3: AC coupled current transformer amplifier.

Adding Cx solves the DC problem, but creates a new one.
We now have a second order filter Ls and Cx.
This filter must have a Q of < 0.5 to prevent it from ringing, which can cause periodic overload with signals close to the amplifier's clipping limit.
For a Beyer 310 001 001 ( Rs = 75 Ω, Ls = 6 H ) this is a minimum value of 9.6 mF for Cx ( see appendix B for calculations ).
Adding a high-pass filter in front of the amplifier can reduce the value of Cx considerably.

INIC loaded transformer amplifier schematic.
Fig.4: INIC loaded transformer amplifier.

Modifying the above circuit to an Inverting Negative Impedance Converter (INIC) can further reduce THD.
In fig.3 there will be some field in the transformer due to Rs. This can be reduced by loading the transformer with a negative resistance.
For R25 = R27, the resistance into node A is - R23.
Setting R23 = Rs will set the field in the transformer to 0, but for stability R23 < Rs.
When selecting R23, the tolerance and temperature coefficient of Rs ( a piece of copper wire ) must be taken into account to assure that the circuit is stable.
For a TR/BV 310 001 001, 51 Ω is a suitable value.
This circuit have similar DC problems (actually worse due to the positive feedback) as the circuit of Fig.2.
Adding Cx to this circuit is not practical as the value will be 470 mF (not a typo - it says millifarads ).

INIC loaded transformer amplifier with servo schematic.
Fig.5: INIC loaded transformer amplifier with servo.

Table 1: Some measurements on the above circuits (on solder-less breadboard).
FigFrequencyInput levelOutput levelTHD+NPrimary THD+N
DC output voltageOutput noise
220 Hz22.01 dBu20.90 dBu0.3%2.harmonic
2100 Hz22.01 dBu20.99 dBu0.012%2.harmonic
21 kHz22.01 dBu21.02 dBu0.0015%2.harmonic
2-40 mV-102.7 dBu
420 Hz22.01 dBu21.10 dBu0.055%2.harmonic
4100 Hz22.01 dBu21.12 dBu0.0036%3.harmonic
41 kHz22.01 dBu21.13 dBu0.0008%3.harmonic
4-300 mV-102.7 dBu
520 Hz22.01 dBu21.10 dBu0.048%2.harmonic
5100 Hz22.01 dBu21.12 dBu0.0031%3.harmonic
51 kHz22.01 dBu21.13 dBu0.0008%3.harmonic
51 mV-102.8 dBu

THD+N and noise measurements are with a 22 Hz to 22 kHz bandwidth.
Component values:
T21: TR/BV 310 001 001.
Signal OP-AMP: NE5532A.
Servo OP-AMP: TL072.
R21, R22: 4.99 kΩ.
R23: 51.1 Ω.
R25, R27: 10 kΩ.
R24: 100 kΩ.
R28: 1 MΩ.
C26: 1 µF.


IA19B schematic.
Fig.6: IA19 schematic ( left channel ).

J21 is a female XLR connector. Neutrik A series and other types with similar footprints can be used.
The 2 nets CH1L and CH2L are brought out on solder pads near the panel and can be connected to the panel with solder tabs. This is useful for connector types where pin 1 and chassis are not internally connected to chassis.
The solder-terminals LI+, LI- and LIG can be installed in stead of the XLR.
R21 and R22 are the input series resistors. These were matched to within 0.1% in the prototypes.
C21..C24 are capacitors to balance the input capacitance in the transformer.
T21 is a universal transformer footprint that fits a Neutrik NTM 1 or Bourns LM-NP-100x series transformers with single, split or center-tapped primary or secondary.
D21 and D22 are input protection diodes. Ideally they should go from the transformer to GND, but the voltage can be so high they start to conduct and cause distortion.
The values of R23 and R24 depends on the transformer type used.
C27 is a DC blocking capacitor for the output. If you can live with some mV DC out, you can short it.
R29 and R30 provides an impedance balanced output.

Component selection.


All electrolytic capacitors are general-purpose types with sufficient voltage rating.
Supply decoupling capacitors C28, C29, C48, C49 ( not shown in fig. 6 ) must be ceramic types.
Low value capacitors C21..C24 must be ceramic NP0 types, preferably with a voltage rating of 100 V or more.
C21 and C23 are through hole types with 2.5 mm pin-spacing.
C22 and C24 are 1206 SMD types as low value through hole capacitors can be difficult to get.
Only one of C21..C24 are needed, see appendix A.
The capacitor values for the prototypes are C23 = 22 pF for the Neutrik transformer and C21 = 4 pF to 5 pF for the Bourns transformer.
C26 is a 5% or 10% film capacitor.


The two diodes are standard signal diodes like 1N4148 or 1N914.


All resistors are 0.25 W..0.6 W 1% metal-film types.
The value of R23 depends on the transformer type used. Use IA19B_Calc.ods in the design files download to calculate it.
I used 48.7 Ω for the Bourns transformer and 75 Ω for the Neutrik transformer.


The board accepts Neutrik NTM 1 transformers or Bourns LM-NP-100x series transformers.
The Neutrik transformers are quite expensive at around €40 from an authorized Neutrik distributor while the Bourns transformers are widely available for around €2.
The prices are from 2018.


U1 is a dual FET OP-AMP in a SOIC-8 case. It must be unity-gain stable.
I have tested the circuit with OPA1678 and TL072.
The OPA2134 and OPA1642 will probably work well, but they are more expensive than the OPA1678.


The input connections can be made with Neutrik A or B series connectors or solder terminals can be used.
If you use a connector similar to the Neutrik connector from another manufacturer, make sure that no metal on the connector shorts to the pads for the input solder-terminals.
The power and output connections can be solder terminals, three 3-pin pin-rows or a 11-pin pin-row. The pin-spacing for the pin-rows is 2.54 mm.


This is the specification for the IA19B prototypes.
Supply is ±15 V from a low-noise lab power supply (<-92 dBu (18 µV) noise in the 22 Hz to 22 kHz bandwidth).
The unit dBu is dB referred to 0.775 V.
More detailed measurements are in the file IA19B_Measurements.ods in the design file download.

Table 2: Specification for IA19B. Unless stated, source impedance is 40 Ω and load impedance 100 kΩ.
TransformersNTM 1LM-NP-1001-B1LLM-NP-1001-B1L
Supply current, no signal, no load8.5 mA8.1 mA7.3 mA
Input impedance, 20 Hz..20 kHz9.4 kΩ..10.1 kΩ9.8 kΩ..10.1 kΩ9.8 kΩ..10.1 kΩ
Voltage gain, 1 kHz, left channel-0.07 dB-0.06 dB-0.08 dB
Voltage gain, 1 kHz, right channel-0.06 dB-0.05 dB-0.07 dB
Linearity, 20  Hz..20 kHz, ref. 1 kHz±0 dB ( Note 1 )+0 dB / -0.15 dB+0 dB / -0.16 dB
Output clip level, 1 kHz, <1% THD+N22.6 dBu ( 10.5 V )22.6 dBu ( 10.5 V )21.4 dBu ( 9.1 V )
THD+N, 1 kHz, 22 Hz..22 kHz BW, 6 dBu ( 1.5 V ) input.0.0007%0.0007%0.0008%
THD+N, 1 kHz, 22 Hz..22 kHz BW, 20 dBu ( 7.7 V ) input.0.0005% ( Note 1 )0.0005% ( Note 1 )0.0006%
THD+N, 20  Hz..20 kHz, 10 Hz..80 kHz BW, 6 dBu ( 1.5 V ) input.<0.02%<0.03%<0.03%
THD+N, 20  Hz..20 kHz, 10 Hz..80 kHz BW, 20 dBu ( 7.7 V ) input.<0.02%<0.03%<0.03%
IMD, SMPTE, 60 Hz / 7 kHz, 4:1, 6 dBu ( 1.5 V ) input.<0.002%<0.002%<0.002%
IMD, SMPTE, 60 Hz / 7 kHz, 4:1, 20 dBu ( 7.7 V ) input.<0.002%<0.002%<0.002%
Cross-talk, 1 kHz-136 dB-100 dB-100 dB
Cross-talk, 20  Hz..20 kHz<-120 dB<-83 dB ( Note 2 )<-83 dB ( Note 2 )
Output noise, 22 Hz..22 kHz, Rs=0 Ω<-107.4 dBu ( 3.3 µV )<-104.8 dBu ( 4.5 µV ) ( Note 3 )<-100.8 dBu ( 7.1 µV )
Output noise, 22 Hz..22 kHz, Rs=1 kΩ<-107.5 dBu ( 3.3 µV )<-104.8 dBu ( 4.5 µV ) ( Note 3 )<-100.1 dBu ( 7.7 µV )
Output noise, CCIR-468 Q-peak, Rs=0 Ω<-96.2 dBu ( 12 µV )<-95.7 dBu ( 13 µV )<-91.1 dBu ( 22 µV )
Output noise, CCIR-468 Q-peak, Rs=1 kΩ<-96.3 dBu ( 12 µV )<-95.7 dBu ( 13 µV )<-91.3 dBu ( 21 µV )
CMRR ref. IEC60268-3, 50 Hz.99 dB93 dB91 dB
CMRR ref. IEC60268-3, 1 kHz.84 dB82 dB80 dB
CMRR ref. IEC60268-3, 20  Hz..20 kHz.>57 dB>70 dB>65 dB
Board size including XLR connectors ( length / width / height )84 mm / 48 mm / 31 mm ( Note 4 )

Note 1:Below the resolution of my audio analyzer.
Note 2:Cross-talk can be increased with a mu-metal shield between the transformers.
Note 3:This can be improved by 3 dB by shielding the transformers.
Note 4:The board is <46 mm wide allowing a channel spacing of 23 mm.


IA19BAA photo. IA19BCA photo.
Fig.7: Photos of mounted boards.

Download IA19B design files.

I have boards available for this project. See the PCBs page.

Download IA19A files. Unpack the zip file into a separate directory (links will not work). PCBs are currently available for this design.

Known Issues / updates.

No known issues.

Appendix A: CMRR adjustment.

For best HF CMRR, the compensation capacitance for the transformer must be trimmed.
The trimmed performance is typically 20..30 dB better than the untrimmed.

CMRR adjustment setup schematic.
Fig.A1: CMRR adjustment setup.

G is a sine-wave generator with a frequency in the range 10 kHz to 20 kHz and 5 V to 10 V output voltage.
M is ideally a meter with a band-pass filter tuned to the generator frequency, but an ordinary AC voltmeter can be used.
The meter must be able to measure at the generator frequency. Its accuracy is not important, but it should be able to display down to 1 mV.
RHP and CHP is a high-pass filter at 5 kHz that allows you to hold your fingers close to the transformer input without upsetting the measurement too much.

You will need a selection of capacitors in the range 1 pF to 22 pF ( the E-6 series is sufficient ).
With power, signal and meter connected turn the PCB upside down and and hold a capacitor ( start with 10 pF ) across C21 or C23.
The meter reading should decrease with the capacitor in one position. If its decrease with the capacitor in the C21 position, the C21 position is used ( and nothing goes in C23 ).
Then find the capacitor that give the lowest meter reading.
If your transformers are from same batch, it is very likely that the same capacitor value will work for both channels.

Appendix B: Some calculations.

Below some calculations for the transformer input stage.
These are also in the IA19B_Calc.ods spreadsheet.

Inverting Negative Impedance Converter (INIC) schematic.
Fig.B1: Inverting Negative Impedance Converter (INIC). Cx is not in IA19.

The values of R23, R25 and R27 must be chosen so R27 / Rs < R25 / R23 for the circuit to be stable.
Rs is a piece of copper wire, so do take its temperature coefficient into account.

Voltage gain:
Av ≈ R27 / ( R21 + R22 + Rp) / n, where Av is the voltage gain from IN to OUT and n is the transformer turns ratio.
Note that the transformer is used in current mode so a voltage step-up transformer is a step-down transformer in this circuit.

Input impedance:
Rin ≈ R21 + R22 + Rp

INIC input impedance ( into node A ):
B is the positive feedback divider ratio, B = R23 / ( R23 + R25 )
RinA = R27 * ( A * B - 1 ) / ( A * B - 1 - A ), where A is the OP-AMP open-loop gain.
For A >> 1:
RinA ≈ - R23 * R27 / R25
For R25 = R27:
RinA ≈ - R23

AC coupling capacitor Cx.
This is often used in this circuit to avoid DC current in the transformer, but the value must be very large to avoid ringing with transient input signals.
The resonance frequency and Q of this circuit is:
f0 = 1 / ( 2 * π * √Ls * Cx )
Q = √Ls * Cx / ( RinA + Rs )
Note that RinA is negative.
To get Q = 0.5 ( value for critical damping ):
Cx = Ls / ( 0.5 * ( RinA + Rs ) )²

Appendix C: Neutrik and Bourns PCB transformers.

Beyer Dynamics series 310 pin-out.
Fig.C1: Neutrik NTM 1 and Bourns LM-NP-100x series pin-out.

NTM 1:
Pin 1 is the core, static ( Faraday ) shield and µ-metal case.
Pins 6 and 8 are the ends of the primary.
Pin 7 can be used for phantom power ( the transformer is designed for microphone input ).
Pin 4 is the start ( innermost ) end of the secondary winding ( I would normally expect the end with the polarity dot to be start-of-winding ).
With a common mode signal on pins 6, 7, 8, the CMRR at 20 kHz is 36 dB with GND on pins 1 and 2 while it is 58 dB with GND on pins 1 and 4.
It would have been nice if Neutrik had included this information in the data-sheet.
Pins 3, 5 and 9 are not connected internally. Grounding these improves CMRR by around 1 dB.

I have only tested the LM-NP-1001-B1L in the IA19B PCB, but there are pads and connections for the other 4 types to work.
Pins 1 and 8 are in phase as shown ( for some reason Bourns do not mention this in the data-sheet ).
The secondary connection is not important. Of 5 samples I tested there is around 1 dB difference in CMRR with pin 5 or pin 8 on GND.

Table D1: Specification for Neutrik NTM 1 and Bourns LM-NP-1001-B1L.
Neutrik NTM 1Bourns LM-NP-1001-B1L
Turns ratio1:1 ±10%1:1 ±2%
Primary DC resistance85 Ω ±3%66 Ω ±10%
Secondary DC resistance105 Ω ±5%66 Ω ±10%
Primary inductance>8 H @ 30 Hz
( 25 H, 21 H )
Min 2.8 H @ 200 Hz
Leakage inductanceNot specified
( 2.0 mH, 1.8 mH )
14 mH
Numbers in italics are measured values.
You can find further specifications for the 2 types in their data sheets, but they are not relevant for this application.

Note that all measurements have been made with the PCB on a table with no electrical shielding at all.
Performance will most likely improve when the boards are mounted in a chassis.


NTM 1:
This is typical of a small audio transformer: around 100 dB @ 50 Hz falling to around 60 dB at 20 kHz.
This was a positive surprise: around 95 dB @ 50 Hz falling to around 70 dB at 20 kHz.

Magnetic shielding

NTM 1:
The transformer is housed in a quite substantial µ-metal case. It is very tolerant of nearby magnetic fields. Do not put it on top of a mains transformer though - the µ-metal will saturate.
This is surprisingly tolerant of magnetic fields for an unshielded transformer. It can be shielded with µ-metal, but unless you have the know-how and tools to work with µ-metal, you will most likely end up with ( very expensive ) sheet iron.
A flat µ-metal shield between the 2 transformers will improve channel separation by 10 dB to 20 dB.
I tried to shield the transformer with plain sheet iron, but distortion will increase when it gets within some mm of the transformer.


NTM 1:
Not much to say: I can not measure it. It is within the 0.01 dB resolution ( not accuracy ) of my audio analyzer.
Flat to 5 kHz and drops 0.05 dB @ 10 kHz, 0.15 dB @ 20 kHz. Pretty amazing for a €2 transformer.

Some general precautions for handling signal-transformers are listed in [1], page 28.

Appendix D: Beyer Dynamics PCB transformers.

Although out of production for many years, these transformers are quite popular and I have had several questions about their specifications.
Part-number: TR/BV 310 abb ccc
310 means a studio grade dual-in-line PCB transformer.

a is winding configuration, where:
0 has a single primary.
1 has a center tapped primary.
2 has split bi-filary primary.
3 has split bi-filary primary connected in the middle (single winding brought out).

Beyer Dynamics series 310 pin-out.
Fig.D1: Beyer Dynamics series 310 pin-out.
The dot marks the start (inner-most end) of the winding.
Pin 8 is static shield and core.
Pins 7, 8 and 9 goes to GND for best CMRR.

bb is turns ratio, where turns ratio is 1:bb.
00 means a special turns ratio (probably a custom manufactured type).

ccc is an type number for the particular transformer:
001 to 020: standard transformer, Rgen = 200 Ω.
021 to 030: special transformer, Rgen = 200 Ω.
031 to 050: standard transformer, Rgen = 600 Ω.
051 to 060: special transformer, Rgen = 600 Ω.
061 to 080: standard transformer, Rgen = 1200 Ω.
081 to 100: special transformer.
101 to 120: special transformer, turns ratio ≤ 1.
251 to 275: standard transformer, Rgen = 200 Ω
276 to 299: standard transformer, Rgen = 500 Ω

Table D1: Some standard types for 200 Ω source impedance.
TR/BV310Rgen = 200 Ω
at Rgen
(open secondary)
30 Hz - 15 kHz ±1 dB
Maximum input level
( THD ≤ 1%, f ≥ 30 Hz )
300 mV
Primary open circuit inductancetyp. 6 H @ 50 Hz
Primary open circuit impedancetyp. 1.885 kΩ @ 50 Hz
Turns ratio1:11:31:51:71:101:151:20
Transformed impedance200 Ω2 kΩ5 kΩ10 kΩ20 kΩ45 kΩ80 kΩ
Primary DC resistance50 Ω50 Ω50 Ω50 Ω85 Ω85 Ω115 Ω
Secondary DC resistance75 Ω580 Ω1.45 kΩ3.45 kΩ4.75 kΩ10.4 kΩ13.5 kΩ
Code number ( bbccc )01001030020500307004100051500620007

Table D2: Some standard types for 600 Ω source impedance.
TR/BV310Rgen = 600 Ω
at Rgen
(open secondary)
30 Hz - 15 kHz ±1 dB
Maximum input level
( THD ≤ 1%, f ≥ 30 Hz )
300 mV
Primary open circuit inductancetyp. 16 H @ 50 Hz
Primary open circuit impedancetyp. 5 kΩ @ 50 Hz
Turns ratio1:11:21:31:51:71:10
Transformed impedance600 Ω2.4 kΩ6 kΩ15 kΩ30 kΩ60 kΩ
Primary DC resistance140 Ω140 Ω140 Ω140 Ω190 Ω190 Ω
Secondary DC resistance190 Ω575 Ω1.45 kΩ3.9 kΩ7.05 kΩ11 kΩ
Code number ( bbccc )010310203203033050340703510036

Table D3: Some standard types for 1200 Ω source impedance.
TR/BV310Rgen = 1200 Ω
at Rgen
(open secondary)
30 Hz - 15 kHz ±1 dB
Maximum input level
( THD ≤ 1%, f ≥ 30 Hz )
300 mV
Primary open circuit inductancetyp. 30 H @ 50 Hz
Primary open circuit impedancetyp. 10 kΩ @ 50 Hz
Turns ratio1:11:21:31:5
Transformed impedance1.2 kΩ5 kΩ12 kΩ30 kΩ
Primary DC resistance310 Ω310 Ω310 Ω475 Ω
Secondary DC resistance430 Ω1.4 kΩ3.75 kΩ8.45 kΩ
Code number ( bbccc )01061020620306305064

These transformers do not like DC.
I tried with a TR/BV 310 001 001 coupled as a voltage transformer with 40 Ω source impedance and 100 kΩ load.
Initial THD+N at 70 mV, 100 Hz was 0.022%, mainly 2. harmonic (there has obviously been DC on this one).
After applying 100 µA from an ohmmeter for a few seconds, THD+N increased to 0.028% still mainly 2. harmonic.
After demagnetizing the transformer, THD+N fell to 0.017%, but this time only 3. harmonic.
Demagnetizing was done by increasing the input level until the transformer went into saturation (ca. 25% THD+N), leaving it there for some seconds and then reducing the level to 0 over 20..30 seconds.
It is possible to demagnetize the transformers while they are mounted in the IA19:
Disconnect the power-supply.
Apply 10 Hz / 6 dBu ( 1.5 V ) to drive the transformer into saturation ( for the TR/BV 310 001 001 ) and reduce the level slowly.
The transformer output waveform can be observed on the top of diodes D22 and D42.

Leakage inductance is not specified for these transformers.
I measured 2 transformers with pins 1, 5, 7, 8, 9 connected to GND.
This was done by putting a 1 µF capacitor in series with the primary and measuring the resonance frequency.
TR/BV 310 001 001 is 2.1 mH and TR/BV 310 001 031 is 4.4 mH

Mechanically these transformers seems very rugged, but internal wiring is down to 25 µm for some types, so handle with care.
Pins 1..8 are gold-plated and very easy to solder.
Pin 9 is raw my-metal and can be difficult to solder. The trick is to take a scalpel and scrape it until it shines just before you solder it.
Some general precautions for handling signal-transformers are listed in [1], page 28.

Appendix E: Some test circuits.

These are some simple circuits to help adjusting and measuring the amplifier.
They can easily be built on strip-board or solder-less breadboard.

Sine-wave generator for CMRR adjustment.
Fig.E1: Simple sine-wave generator for CMRR adjustment (only simulated - not tested).

Simple voltmeter for CMRR adjustment.
Fig.E2: Simple voltmeter for CMRR adjustment (only simulated - not tested).
Power supply connections as in Fig E1.

Switch for CMRR measurement.
Fig.E3: Switch for CMRR measurement. See [1], page 24 for test setup.
For testing bare IA19 boards, the ground of the generator must be connected to IA19 circuit ground as pin 1 of the XLR is separate from circuit ground.


[1] Audio Transformers by Bill Whitlock.

Copyright and disclaimer.

Copyright Notice.
This web-page, including but not limited to all text, drawings and photos, is the intellectual property of Poul Petersen, and is Copyright ©.
Reproduction or re-publication by any means whatsoever is strictly prohibited under International Copyright laws.
The author grants the reader the right to use this information for personal use only.
Any commercial use is prohibited without express written authorization from Poul Petersen.

The information is provided on an "as-is" basis and is believed to be correct, however any use of the information is your own responsibility.

This web-site may contain links to web-sites outside Poul Petersen domain ( www.poulpetersen.dk ).
Poul Petersen has no control over and assumes no responsibility for the content of any web-site outside Poul Petersen own domain.

Poul Petersen does not use cookies to "enhance your experience" on this website.

Poul Petersen, C/Faya 14, 35120 Arguineguín, Las Palmas, Spain.
E-mail: diy@poulpetersen.dk

Poul Petersen DIY index

Poul Petersen notes index

Copyright © Poul Petersen 2018. Last update: 20191029. Valid HTML!