A time locked, low level calibrator

A time locked, low level calibrator

616 C W DARROW AND H F SMITH a resulting negative pulse to the base of transistor X4 If during this negatwe pulse there is closing of contact B, a p...

206KB Sizes 3 Downloads 37 Views

616

C W DARROW AND H F SMITH

a resulting negative pulse to the base of transistor X4 If during this negatwe pulse there is closing of contact B, a positive voltage is apphed to the emitter of X4 with resulting current flow, and a deflection of the penwrlter (I) Slmdar but opposite effects will occur when EEG waves occur in opposite sequence and contact A succeeds contact B Opening of contact B w)th recharging of condenser Cz will result in a negative pulse to the base of transistor X3 preparing it for conduct)on If, during this negative pulse, contact A is then closed, a pomtive charge on the emitter of X3 will result In gating of X3 and an appropriate pip on the penwrlter (2) Results of the operation of ttus scoring system m parallel with a record of the phase indicator are shown during the beating of two generator-produced frequencies in Fig 2 In Fig 3 is an excerpt showing the apphca-

tton of the method during EEG recording Seven phase scorers are currently in use recording the inter-area time relationships between EEGs from six electrode placemeats SUMMARY A transistor-operated instrument providing continuous automatic scoring of inter-area EEG phase relations is described REFERENCE DARROW, C W , WILCOTT, R C , SIEGEL,A B , STROUP, M and AARONS, L Instrumental evaluation of EEG phase relationships Electroenceph chn Neurophyslol, 1956, 8 333-336

Reference DARROW, C W and SMITH, H F An instrument for automatic scoring of EEG phase relationships Electroenceph chn Neurophystol , 1964, 16 614-616

A TIME LOCKED, LOW LEVEL CALIBRATOR 1 JOHN W EMDE

Division of Medical Electromcs, College of Medlcme, State Umverstty of Iowa, Iowa City, Iowa (U S A ) (Accepted for publication July 17, 1963) When recording electrophysiologlcal data, it is often convenient to have a cahbratIon signal superimposed on the record to provide a known voltage reference If an averaging technique is in use, this calibration pulse must 1 Supported in part by USPHS grant number MY2635

be time locked to the stimulus necessitating some form of electronic waveform generation The calibrator described was designed primarily for use when averaging evoked responses in conjunction with the CAT model 400A, however, the author feels it may have application in other related areas as well Electrodes

VI

V2o

V2b

V3o

I RI

Cl

Input ¢R5

R2 ! R3*

F

i"'l IRI3

V3b

R4 - - -

,J

'+225

R2~

TI

R 15~ Jv~

R R6

Amp Grid

Amplifier

Fig 1 Schematic

Electroenceph chn NeurophysIol, 1964, 16 616-618

A TIME LOCKED, LOW LEVEL CALIBRATOR

~

617 Secondory Leods

I

p Fig 2 Transformer TI To be of optmaum value, the calibration stgual should be injected in series with the subject leads Thus changes m amplifier gain will affect both the evoked response and calibration pulse in the same manner A translstor~zed calibrator, carefully isolated from ground to minimize stray capacitance and triggered by a flaslung neon lamp, has been tested and discarded as useless when inserted in series with the subject leads After consideration o f the dttficulty of maintaining a balanced amphfier input, as seen from the calibrator through the subject leads, it becomes apparent that it is desirable to generate a microvolt level calibration pulse rather than attenuate a l'ugh level pulse As the cahbrator described in this article generates slguals in the microvolt range, it is possible to inject the cahbratmn signal in series with the subject leads with no apparent effect on either the evoked response or the calibration signal due to stray capacitance to ground The total stray capacitance from the transformer secondary and divider switch to ground is of the order of 125 # p F If a ramp function of current is applied to the prunary of a transformer, a step function of voltage will be produced on the secondary wtuch is proportional to the rate of change of current m the primary The constant of proportionality is dependent on the characteristics of the transformer core and on the turns ratio As can easily be seen, the current in the primary of the transformer cannot increase indefinitely, thus a trapezoidal waveform ts utlhzed to prowde the ramp runup, hold the current constant for a predetermined time, then reset This will produce two rectangular pulses on the secondary, one positive going and one negatwe going As destgued, this cahbrator utilizes only the first pulse The second pulse is delayed to fall beyond the recorded analysts time The schematic is shown m Fig 1 When a positive trigger pulse is applied, phantastron V1 starts its rundown The screen waveform is amphfied, reverted, and squared by V2 and is applied to the integrating amphfier V3 The output of V3astarts at approximately ground potential and rises at a constant rate, determined by the output of V2b, R15 and C3, to approximately 25 V where it is caught by CR2 At the end of the phantastron rundown, the screen voltage of VI falls, causing the integrating amplifier to reset The exact endpoints of the runup and rundown are determined by CR1, CR2 and CR3, but as varmtmn of the endpotnts wdl change only the cah-

bratlon pulse duration, not its amplitude, small variations here are not considered troublesome. The trapezoidal waveform is fed to transformer T1 via resistor R23 Resistor R23 is selected to provide a rate o f change of current of 5 5 mA/sec in the primary of T1 dttrmg the ramp runup The secondary ofT1 Is brought out to a 40f~ voltago divider to provide convenient selection of calibration amplitude With the rate of change of ctmrent In the primary of TI set at 5 5 mA/sec, approximately 100 #V will appear across the divider. Final adjustment o f output amplitude may be made by setting R11 To minimize stray pickup, care should be exercised in the placement of R23, also, wiring from the cathode of V3 to R23 and from R23 to transformer TI should be sluelded For convenience o f operation, transformer TI and the divider network are located in a separate box next to the pre-amphfier and connected to the waveform generator chassis by means of a coaxial cable All components with the exception of TI are readily available from electronics supply houses Details oftransformerT1 are given in Fig 2 To simplify construction of the transformer, a torpid choke (Triad ET-1000) is used as the primary The secondary bobbin is machined from brass stock to the dtmensmns shown As can be seen, the crosssection of the torpid is not ctreular so it is necessary to slightly flatten the bobbin by applying pressure at P-P The bobbin IS then cut with a thin blade saw at A-A and re-assembled over the torpid as shown Note that a strip of teflon is placed between the bobbin and the torpid as a heat shield to protect the torpid during the soldering operatmn. Note also that the bobbin is soldered at one cut only and that a small air gap is left at the other cut The secondary, 100 turns of number 26 Nyclad wire, ts now wound on the bobbin and the leads brought out at the air gap A strip of brass shim stock is then soldered over the bobbin to complete the shield To prevent abrastun, the secondary leads are positioned in the center o f the air gap and held in place by a drop o f epoxy The values given for R15, R23, C3 and CR3 will provide a pulse wluch is 5 per cent of the analys~s t~me as hsted in the table of components R3 and C2 determine the time between the calibration pulse and the reset pulse and are selected to provide a delay approxanately 20 per cent greater than the analys~s time These values could, of course, be selected to produce any pulse width and delay desired w~thm the limits of the system If desired, a second

Electroenceph chn Neurophysml, 1964, 16 616-618

618

J. W EMDE TABLE OF COMPONENTS

Unless othervose spectfied, all resistors are 4- 10~o, 0 5 W R1 R2 R3* R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15* R16 R17 R18 RI9 R20 R21

27k~,lW 1 M~ see chart 15 k ~ 22k.Q, l W 15k~ 47k~ 390 kf~ 1 MD 1000 kD-I- 1% 50 kf2 Potentaometer 432 k ~ 4- 1% kD 499 k.Q 4- 1% 100 k ~ see chart 68 k ~ 100kf2 50 k12 Potentlometer 470 f2 1 M~ 100 k ~

R22 R23" R24 R25 R26 R27 R28 R29 R30 C1 C2" C3" CR1 CR2 CR3 CR4 CR5 VI V2 V3 T1

1M[2 see chart 20 fl 1~o 10 [2 1% 6[21~ 2 f ) 1% 1~1~o 1 ~2 1 ~ 100 kD

0 001/~F see chart see chart 1N457 1N457 1N720 1N738 1N457 6AS6 12AX7 6U8 see Fig 2

AnalSwRched components ys~s time R3 C2"* R15 C3"* (msec) 4- I0°/o (#F) -4- 1~e (/~F) (M~2) (k~) 62 5 125 250 500 I000 2000 4000

15 27 68 12 27 47 10

0 15 0 15 0 15 0 75 0 75 0 75 0 75

97 6 196 392 51 1 100 200 402

winding may be added to TI to provide a dual channel cahbraUon This has been done on those calibrators currently m use and has proved to be sattsfactory Two windings wdl fit easily m the b o b b m as shown and no apparent cross couphng has occurred A typical record ~s shown m Fig 3 Note that the trigger pulse has been delayed beyond the start of the sweep This Is not entirely necessary but it does prowde a convement method of estabhshlng the cahbratlon basehne Prehmmary tests indicate that no dlstorUon is introduced into the record by the cahbrator and amphtude stability appears to be good

0 056 0 056 0 056 0.3 03 03 03

~100 ms [--- :Stimulus

R23 4- I~o (k,O) 1400 698 348 174 86 6 43 2 22 1

* Switched component, see chart (second part of this table) ** Padding may be required

I ~(~

Fig 3 A t y p t c a l record showing the sum of 100 responses to a photlc StLmulus AS can be seen, the cahbratlon pulse provides a convement amphtude reference The stmlulus marker is followed by a sea~nd marker to provide a Ume reference

SUMMARY A trtggered cahbrator providing pulses m the range of 5 to 100 pV is described The cahbraUon pulse may be added to the blologtcal signal at the pre-amphfier input with no detrimental effects to either signal

Reference EMDE, J W A umelocked, low level cahbrator Electroenceph chn Neurophyswl, 1964, 16 616-618